The starBase database (version 3.0; http://starbase.sysu.edu.cn/index.php/) was used to predict the binding site of circ_0067934 and miR-5047, miR-1301-3p, miR-670-5p, miR-345-3p, miR-545-3p and miR-3605-5p. In addition, the binding site of miR-1301-3p and KIF23 was predicted using the starBase database.

The GES-1 normal gastric mucosal cell line, and the GC cell lines, AGS and MKN-45 (China Infrastructure of Cell Line Resources, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences), were cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) containing 11 mM glucose, 10% (v/v) FBS (cat. no. 10099141; Gibco; Thermo Fisher Scientific, Inc.), 50 µM β-mercaptoethanol (cat. no. M3148; MilliporeSigma) and 10 mM HEPES (cat. no. H1090; Beijing Solarbio Science & Technology Co., Ltd.) at 37°C with 5% CO₂.

Short hairpin RNA (shRNA/sh)-circ_0067934-1/2 and corresponding negative control (NC) shRNA (sh-NC), miR-1301-3p mimic and its NC (mimic-NC), a pcDNA3.1 expression vector containing full-length human KIF23 (Ov-KIF23) and corresponding NCs (Ov-NC) were obtained from Shanghai GenePharma Co., Ltd. The sequence for the miR-1301-3p mimic was 5′-UUGCAGCUGCCUGGGAGUGACUUC-3′. The NC shRNA sequence was 5′-UCACAACCUCCUAGAAAGAGUAGA-3′. The AGS cell suspension was inoculated into a 6-well plate and cultured in an incubator with 5% CO₂ at 37°C for 24 h. The cultured cells were transfected with 100 nM of these recombinants using 2.5 µl/ml Lipofectamine^® 2000 transfection reagent (Thermo Fisher Scientific, Inc.) at 37°C for 48 h in accordance with the manufacturer's instructions. After transfection, the transfection efficacy of the cells was detected via reverse transcription-quantitative (RT-q)PCR.

AGS cells from each group were collected after transfection for 48 h. After being counted, cells (1×10³ cells/well) were seeded into a 96-well cell culture plate and cultured in an incubator with 5% CO₂ at 37°C. After 24, 48 and 72 h, 10 µl Cell Counting Kit-8 (CCK-8) reagent (Beyotime Institute of Biotechnology) was added to each well in accordance with the manufacturer's instructions. Then, the cells were incubated at 37°C in the dark for 2 h. The absorbance of each well was measured at 450 nm with a microplate reader (Bio-Rad Laboratories, Inc.). Each experiment was repeated ≥3 times. In total, six duplicate wells were examined in each group.

AGS cells were seeded into 6-well plates (500 cells/well) and incubated at room temperature overnight. EdU solution (Beyotime Institute of Biotechnology) was added and the cells were incubated for 4 h at 37°C. After the working fluid was removed, the cells were fixed with 4% paraformaldehyde for 15 min at room temperature. The cells were then permeated with 0.5% Triton X-100 for 15 min. After the addition of Click reaction solution, the cells were incubated in the dark for 30 min at 37°C and imaged under a fluorescence microscope (Leica Microsystems GmbH; magnification, ×200).

AGS cells were harvested and total protein was extracted using RIPA lysis buffer (Beyotime Institute of Biotechnology). Total protein was quantified using a protein concentration determination kit (cat. no. P0012; Beyotime Institute of Biotechnology). Proteins (30 µg/lane) were separated via SDS-PAGE (15%) and were subsequently transferred onto PVDF membranes (MilliporeSigma). After blocking with 5% BSA (Beyotime Institute of Biotechnology) at room temperature for 2 h, the membranes were then incubated with the following primary antibodies (Abcam) at 4°C overnight: Anti-Ki67 (cat. no. ab15580; 1:1,000), anti-proliferating cell nuclear antigen (anti-PCNA; cat. no. ab92552; 1:1,000), anti-MMP2 (cat. no. ab92536; 1:1,000), anti-MMP9 (cat. no. ab76003; 1:1,000) and anti-KIF23 (cat. no. ab174304; 1:1,000). After the incubation with the primary antibodies, the membranes were washed with TBS containing Tween-20 (0.1%) and incubated at room temperature for 1.5 h with HRP-conjugated secondary antibodies [Goat Anti-Mouse IgG(H+L), 1:2,000, cat. no. SA00001-1; Goat Anti-Rabbit IgG(H+L), 1:2,000, cat. no. SA00001-2; ProteinTech Group, Inc.]. Protein bands were visualized using the Odyssey Western Blot Analysis system (LI-COR Biosciences) and semi-quantified using ImageJ software, version 7.6.5 (National Institutes of Health).

The invasive ability of the cells was examined using a Transwell invasion assay. The Transwell chambers (Costar; Corning, Inc.) were first coated with 0.1 ml Matrigel (Becton-Dickinson and Company) at 37°C for 1 h. AGS cells (5×10⁵ cells/well) were collected and suspended at 2×10⁵ cells/ml in serum-free DMEM. Cell suspensions were placed into the upper compartment of a Transwell chamber and cultured in 5% CO₂ at 37°C. Medium supplemented with 10% FBS was added into the lower compartment of the Transwell chamber. After 24 h, the non-invaded cells on the upper face of the Transwell membrane were wiped off with a cotton swab. The cells invading to the lower surface of the filter were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet solution for 30 min at room temperature. Images were captured using a light microscope at ×100 magnification.

AGS cells were cultured in 6-well plates (6×10⁴ cells/well) and incubated to 80–90% confluence in RPMI-1640 medium with 10% FBS at 37°C (17). A linear scratch was created in the cell monolayer using a 200-µl pipette tip. The cells were subsequently cultured under standard conditions in RPMI-1640 medium containing 2% FBS (Procell Life Science & Technology Co., Ltd.) for 48 h at 37°C in 5% CO₂. The area occupied by cells migrating into the scratch was evaluated using an inverted microscope at ×100 magnification. The migration rate was calculated based on the formula: (Wound width at 0 h - wound width at 24 h)/wound width at 0 h × 100%. In total, five fields were randomly selected for analysis in each well.

AGS cells were lysed and centrifuged for 5 min at 500 × g at 4°C. Then, the cytoplasmic components retained in the supernatant were collected. The precipitated nuclear components were further cleaved. The collected cytoplasmic components and nuclear lysates were mixed with 2X lysis/binding solution and then pumped through a filter tube. The extracted cytoplasmic and nuclear RNA were reverse-transcribed, and the expression level of circ_0067934 was detected via RT-qPCR. U6 was the positive control for the detection of nuclear RNA expression, and GAPDH was the positive control for the detection of cytoplasmic RNA expression (18).

The circ_0067934 3′-UTR, containing wild-type (WT) or mutant (MUT) target sites for miR-1301-3p, was amplified via PCR and inserted into a pGL3 vector (Promega Corporation) to form the reporter vector circ_0067934-WT or circ_0067934-MUT. Then, the DNA sequence of KIF23 3′-UTR that included the putative binding sites of miR-1301-3p was subcloned into a pGL3 vector (Promega Corporation) to create a luciferase plasmid containing the WT sequence of KIF23 3′-UTR. The miR-1301-3p binding sites were mutated, and the MUT KIF23 3′-UTR sequence was subcloned into the pGL3 vector so as to create a mutant reporter vector. AGS cells were co-transfected with luciferase reporter vectors by using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) for 48 h. At 48 h after transfection, the relative luciferase activities were measured by using a Dual-Luciferase Reporter Assay (Promega Corporation) and luciferase activities were normalized against Renilla luciferase.

Total RNA from cell samples was extracted using TRIzol^® reagent (Thermo Fisher Scientific, Inc.) and reverse-transcribed into cDNA using a commercial RevertAid™ cDNA Synthesis kit (Takara Bio, Inc.) at 42°C for ~1 h and 90°C for 5 min. The SYBR Premix Ex Taq™ II kit (Thermo Fisher Scientific, Inc.) was applied for qPCR. The PCR reaction mixture contained 3 mM MgCl₂, 0.5 µM forward and reverse primers, 2 µl SYBR Green PCR master mix and 2 µl cDNA on a QuantStudio 3 Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The primers used are shown in Table I. The thermocycling conditions of the qPCR were as follows: Initial denaturation for 5 min at 95°C, followed by 40 cycles for 30 sec at 95°C, annealing at 65°C for 45 sec and extension at 72°C for 1 min. A final extension step at 72°C for 7 min was performed in each PCR assay. The results were measured using the 2^−ΔΔCq method (19). U6 served as an endogenous control for miRNAs and GAPDH was used as an endogenous control for circ_0067934 and KIF23.

Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software, Inc.). All data are presented as the mean ± SD of three independent experiments. Statistical differences between two groups were determined using an unpaired two-tailed Student's t-test, while one-way followed by Tukey's post hoc test was used to analyze data in >2 groups. P<0.05 was considered to indicate a statistically significant difference.

To investigate the role of circ_0067934 in GC cells, the expression level of circ_0067934 in GC cells was firstly detected via RT-qPCR. The results demonstrated that the expression level of circ_0067934 was upregulated in the GC cell lines, AGS and MKN-45, compared with the normal gastric mucosa cells, GES-1 (Fig. 1A). AGS cells showed the highest expression of circ_0067934, so these cells were selected for subsequent experiments.

Cells were then transfected with sh-circ_0067934-1/2 to knock down the expression of circ_0067934 (Fig. 1B). Compared with the sh-NC group, the cell proliferation ability of the sh-circ_0067934-1 group was significantly decreased (Fig. 1C and D). In addition, the results of western blotting revealed that the expression levels of Ki67 and PCNA were downregulated following transfection with sh-circ_0067934 (Fig. 1E).

As shown in Fig. 2A-C, following transfection with sh-circ_0067934, the migratory and invasive abilities of AGS cells were decreased compared with the control and sh-NC groups. Additionally, western blotting revealed that the expression levels of the migration-related proteins, MMP2 and MMP9, were significantly downregulated compared with those in the control and sh-NC groups (Fig. 2D), indicating that circ_0067934 silencing inhibited AGS cell migration and invasion.

Next, the expression level of circ_0067934 in the cytoplasm and nucleus of GC cells was examined via RT-qPCR. The results demonstrated that circ_0067934 was mainly present in the cytoplasm of AGS cells (Fig. 3A). The binding sequences of circ_0067934 with miR-5047, miR-1301-3p, miR-670-5p, miR-345-3p, miR-545-3p and miR-3605-5p were predicted using the starBase database (Fig. 3B). Moreover, the RT-qPCR results demonstrated that the expression levels of miR-1301-3p and miR-345-3p were significantly decreased in GC cells compared with control cells (Fig. 3C-H). As shown in Fig. 3I and J, transfection with sh-circ_0067934 promoted the expression levels of miR-1301-3p more significantly than miR-345-3p. Thus, miR-1301-3p was selected for the subsequent experiments.

To determine the effect of miR-1301-3p on AGS cells, miR-1301-3p overexpression vector was transfected into AGS cells (Fig. 3K). Furthermore, the luciferase vectors containing the WT and MUT sequences (shown in red letters) of circ-0067934 were constructed (Fig. 3L) and the results indicated that miR-1301-3p was a direct target of circ_0067934 (Fig. 3M).

As presented in Fig. 4A, the optical density (OD) value of miR-1301-3p mimic-transfected cells was significantly lower compared with that of the mimic NC and control groups. Moreover, EdU staining revealed that transfection with miR-1301-3p mimic markedly inhibited cell proliferation (Fig. 4B). Furthermore, the protein expression levels of Ki67 and PCNA were notably suppressed by miR-1301-3p mimic in comparison with those in the NC and control groups (Fig. 4C).

To further investigate the role of miR-1301-3p in AGS cells, wound healing and Transwell assays were conducted to measure cell migration and invasion. As shown in Fig. 5A-C, the cell migratory and invasive abilities were weakened following transfection with miR-1301-3p mimic compared with that in the mimic NC group. In addition, according to the results of the western blotting assay, the expression levels of the migration-related proteins, MMP2 and MMP9, were significantly decreased following transfection with miR-1301-3p mimic compared with the mimic NC or control groups (Fig. 5D).

The starBase database predicted the targeted binding of miR-1301-3p to KIF23 (Fig. 6A), and luciferase reporter assay verified the association between miR-1301-3p and KIF23 (Fig. 6B). Moreover, it was found that KIF23 expression was significantly upregulated in GC cells lines compared with the normal gastric mucosa cells (Fig. 6C). The results of western blotting identified that the expression of KIF23 was decreased following transfection with miR-1301-3p mimic compared with the mimic NC group (Fig. 6D). These results indicated that miR-1301-3p targets KIF23 in AGS cells.

To determine the role of KIF23 in GC cells, the miR-1301-3p inhibitor and KIF23 overexpression vector were transfected into AGS cells (Fig. 7A-C). The level of cell proliferation was detected using a CCK-8 assay and the results revealed that the OD value was significantly enhanced following transfection with miR-1301-3p inhibitor or Ov-KIF23 in circ_0067934-silenced cells compared with their NC groups (Fig. 7D). In addition, compared with the sh-circ_0067934 + inhibitor-NC group and sh-circ_0067934 + Ov-NC group, miR-1301-3p inhibitor and Ov-KIF23 markedly increased the percentage of living cells (Fig. 7E). Consistently, the western blotting results demonstrated that miR-1301-3p silencing and KIF23 overexpression increased the protein expression levels of Ki67 and PCNA in cells transfected with sh-circ_0067934 (Fig. 7F).

As shown in Fig. 8A-C, KIF23 overexpression or miR-1301-3p knockdown promoted cell migration and invasion compared with the sh-circ_0067934-1 + Ov-NC and sh-circ_0067934 + inhibitor-NC groups. In addition, the western blotting results demonstrated that the protein expression levels of MMP2 and MMP9 were significantly increased following transfection with miR-1301-3p inhibitor or Ov-KIF23 in circ_0067934-silenced cells compared with their corresponding NC (Fig. 8D).