In total, 20 patients with recently diagnosed (and previously untreated) CC at the Department of Obstetrics and Gynaecology of The First Affiliated Hospital of Soochow University from July 2016 to August 2017 were selected in a single-centre study and had a median age of 51 years (range, 39–75). The CC tissues and tumour-adjacent tissues were collected from the selected patients following diagnostic evaluation. Written informed consent was obtained from each patient prior to the experiment. The protocols were also approved by the local ethics committee of The First Affiliated Hospital of Soochow University. All the clinical samples collected in the present study are squamous cell carcinoma.

Human CC cell lines (HeLa-S3 and C-33A), immortalized cervical epithelial cell line (H8) and human embryonic kidney cells (293T) were obtained from the Cell Resource Center of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM (HyClone, Cytiva) containing 10% foetal bovine serum (Gibco, Thermo Fisher Scientific) and 100 U/ml penicillin/streptomycin (HyClone, Cytiva) in a 37°C constant incubator with 95% air and 5% CO₂.

Total RNA of tissues and cell lines was extracted using the RNeasy Plus Universal Tissue kit (Sangon Biotech Co.) based on the manufacturer's instructions. miRNAs from CC cell lines were extracted using the miRNeasy Mini Kit (Sangon Biotech Co). PCR primers for 25 potentially dysregulated miRNAs and U6 were synthesized by Guangzhou RiboBio.

Primers for CXCL8 and GAPDH were synthesized by Takara (Sangon Biotech Co.) (Table I). cDNA was constructed using the PrimeScript RT reagent kit (Takara Biotechnology Co., Ltd.). Real-time PCR was performed using SYBR premix Ex Taq II (Takara Biotechnology Co., Ltd.), and fluorescence intensity was detected in a Light Cycler 480 system (Roche Applied Science) using the following thermocycling conditions: 95°C for 10 min, 40 cycles of 95°C for 15 sec and 60°C for 1 min. GAPDH was used as the internal control. The comparative Cq (^ΔΔCq) method (19) was used to calculate the fold change for each miRNA/mRNA. U6 and GAPDH were treated as internal controls. The following formulas were used to calculate the expression fold changes (X): ΔCq=Cq (target miRNA/mRNA)-Cq (U6/GAPDH), ΔΔCq=ΔCq (target miRNAs/mRNA)-ΔCq (target miRNAs/mRNA), X=2^−ΔΔCq.

Human miR-302c-3p and miR-520a-3p mimics as well as scrambled miR-302c-3p mimics, miR-520a-3p mimics and small interfering RNAs were obtained from Guangzhou RiboBio. A CXCL8 eukaryotic expression vector (CXCL8/pcDNA3.1) was generated by inserting the open reading frame of CXCL8 into pcDNA3.1, using the empty vector as a negative control. For luciferase reporters, the 3′-UTR of CXCL8 was amplified from human genomic DNA before the miR-302c-3p and miR-520a-3p binding sites within the CXCL8 3′-UTR were mutated to remove complementarity. The mutant or wild-type 3′-UTR of CXCL8 were cloned into the psiCHECK-2 luciferase vector (cat. no. 78260; Promega Corporation). All plasmids were confirmed by sequencing.

Cell counting kit (CCK-8, Merck KGaA) and ethynyl-2-deoxyuridine (EdU, Guangzhou RiboBio) staining were both utilised to measure the proliferation ability of each cell line under different conditions. For the CCK-8 assay, 5×10³ cells in 100 µl were transferred to each well of a 96-well plate. Cells were cultured at 37°C before CCK-8 solution (10 µl) was added to each well 24, 48, 72, and 96 h later. Subsequently, the cells were incubated at 37°C for an additional 4 h before OD values (450 nm) were measured using a microplate reader (Bio-Rad Laboratories, Inc.). For the EdU assay, the cells were transfected with miR-302 and miR-520 mimic or inhibitor (100 nM) using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) and incubated with 100 µl of 50 µM EdU per well for 2 h at 37°C. Cells were visualized by fluorescence microscopy and the ratio of EdU/DAPI was calculated to analyse the EdU assay. Experiments were repeated at least three times.

Cell apoptosis was assessed with the Annexin V-FITC apoptosis detection kit (Elabscience). Flow cytometry was performed with FACSCalibur (BD Biosciences) and all experiments were performed at least three times.

The potential direct target genes of miR-302c-3p and miR-520a-3p were predicted by miRTarBase (mirtarbase.mbc.nctu.edu.tw/php/index.php). For the luciferase reporter assays, 293T cells were seeded onto 24-well plates before miR-302c-3p or miR-520a-3p mimics (100 nM) and wild-type or mut-type CXCL8 3′-UTR (1.5 µg) were co-transfected into 293T cells using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). After transfection for 48 h, cells were collected and lysed for luciferase assays. The intensity of luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega Corporation). Renilla luciferase was used for normalization. Each experiment was performed at least three times.

Whole-cell lysates were prepared using RIPA buffer (Beyotime Institute of Biotechnology). Extracts were mixed with SDS-PAGE loading buffer and denatured by boiling for 5 min. Equal amounts of protein (~30 µg) were separated on 10% SDS-PAGE gels and transferred onto polyvinylidene fluoride (PVDF) membranes (Bio-Rad Laboratories, Inc.). Blots were blocked using 5% bovine serum albumin (BSA, Gibco; Thermo Fisher Scientific, Inc.) for 2 h and incubated overnight at 4°C with primary antibodies against CXCL8 (1:1,000; cat. no. ab154390; Abcam) and β-actin (1:5,000; cat. no. 4970S; Cell Signaling Technology, Inc.). PVDF membranes were washed with Tris-buffered saline (TBS) supplemented with 0.05% Tween-20 before incubation with HRP-conjugated secondary antibody (1:2,000; cat. no. 7074S; Cell Signaling Technology, Inc.) for 2 h at room temperature. Signals were visualized using enhanced chemiluminescence reagents (Thermo Fisher Scientific, Inc.). Densitometric analysis of the blot bands was performed using ImageJ 1.48 software, and the intensity values were normalized against the values of the β-actin loading control. All experiments were performed at least three times.

Data were analysed using SPSS software (version 19; IBM Corp.). Continuous data sets are expressed as the mean ± SEM. Comparisons between two groups were analysed using t-tests. One-way ANOVA was used for comparison of three or more groups, with additional post-hoc Tukey's tests to account for multiple comparisons. Statistical analysis of categorical variables was performed using the χ² test. Correlations between miR-302c-3p and miR-520a-3p expression levels and CC profile and between miR-302c-3p, miR-520a-3p and CXCL8 expression levels were determined by Pearson's linear correlation analysis. P<0.05 was considered to indicate a statistically significant difference.

To investigate the potentially dysregulated miRNAs between clinical CC tissues and tumour-adjacent tissues, RT-qPCR was performed to measure the expression of 25 mature miRNA sequences in both CC tissues and tumour-adjacent tissues (Fig. 1A). Two highly downregulated miRNAs, miR-302c-3p and miR-520a-3p, were notable due to their tumour-suppressing roles in other cancer cells (7,18). Furthermore, the expression of miR-302c-3p and miR-520a-3p was examined in 20 pairs of clinical CC tissues and tumour-adjacent normal tissues. The data showed that miR-302c-3p and miR-520a-3p were downregulated in CC tissues compared with adjacent tissues (Fig. 1B and C; ***P<0.001, **P<0.01). Pearson's linear correlation analysis revealed that miR-302c-3p and miR-520a-3p expression levels in adjacent normal tissues were positively correlated with those in normal tissues (Fig. 1D and E; P=0.0051, P=0.0025). The expression levels of miR-302c-3p and miR-520a-3p were measured in two CC cell lines (HeLa-S3 and C-33A), as well as in the immortalized cervical epithelial cell line H8. It was found that miR-302c-3p and miR-520a-3p were both downregulated in HeLa-S3 and C-33A cells (Fig. 1F-G; *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001). These results demonstrate that the expressions of miR-302c-3p and miR-520a-3p are downregulated in both CC tissues and cell lines.

miR-302c-3p is reported to inhibit the growth of hepatocellular carcinoma (20), while miR-520a-3p inhibits the proliferation of non-small cell lung cancer cells (18). To investigate whether miR-302c-3p or miR-520a-3p is involved in the proliferation of CC, CCK-8 and EdU assays were performed using cells transfected with mimics or inhibitors of miR-302c-3p and miR-520a-3p. The transfection efficiency of both overexpression and knockdown of miR-302c-3p and miR-520a-3p were verified using RT-qPCR (Figs. 2A and S1A; ****P<0.0001). A significant decrease in cell proliferation was observed in both HeLa-S3 and C-33A cells transfected with miR-302c-3p/miR-520a-3p mimics (Fig. 2B; *P<0.05, **P<0.01), whereas the cells treated with miR-302c-3p and miR-520a-3p inhibitors had increased proliferation (Fig. S1B; **P<0.01) compared with the control groups (blank cells were treated with an equivalent volume of PBS). Co-transfection of miR-302c-3p and miR-520a-3p mimics led to cooperative suppression of cell proliferation, which was greater than that of cells transfected with miR-302c-3p or miR-520a-3p alone. Results consistent with these observations were also obtained from the EdU staining assay (Figs. 2C and S1C; **P<0.01, ****P<0.0001, *P<0.05). Tumour tissue growth requires the rate of cell proliferation to exceed that of cell death; thus, the apoptosis-associated functions of miR-302c-3p and miR-520a-3p were detected using flow cytometry. As expected, both HeLa-S3 and C-33A cells (Fig. 2D; **P<0.01, ****P<0.0001) transfected with miR-302c-3p/miR-520a-3p mimics showed significant increases in cell apoptosis, whereas the cells treated with miR-302c-3p/miR-520a-3p inhibitor showed decreases in cell apoptosis (Fig. S1D; **P<0.01, ****P<0.0001 vs. control). These results demonstrated that miR-302c-3p and miR-520a-3p can suppress the growth and promote the apoptosis of CC cells in an independent or cooperative manner.

To understand the proliferation-suppressive role of miR-302c-3p and miR-520a-3p, the potential target genes of miR-302c-3p and miR-520a-3p were predicted by using miRTarBase. Among the candidates, CXCL8 has been reported to be upregulated in various types of carcinomas (21,22) and contributes to cancer cell proliferation (23,24). Therefore, CXCL8 was selected for further evaluation and RT-qPCR demonstrated that CC tissues had much higher CXCL8 expression than tumour-adjacent tissues (Fig. 3A; ****P<0.0001). Pearson's linear correlation analysis found that CXCL8 was negatively correlated with miR-302c-3p and miR-520a-3p (Fig. 3B; P=0.0002, P=0.0065). Moreover, RT-qPCR measured the relative expressions of CXCL8 and miR-302c-3p/miR-520a-3p, which showed that high expression of miR-302c-3p or miR-520a-3p corresponds to low expression of CXCL8 (Fig. S2A; **P<0.01). RT-qPCR and western blot analysis revealed that the mRNA and protein expression levels of CXCL8 were increased in HeLa-S3 and C-33A cells compared with H8 cells (Fig. 3C; **P<0.01, ***P<0.001). To investigate whether miR-302c-3p and miR-520a-3p regulate CXCL8 expression in CC cells, HeLa-S3 and C-33A cells were transfected with miR-302c-3p and miR-520a-3p mimics before measurement of CXCL8 expression. CXCL8 was downregulated in response to transfection of miR-302c-3p and miR-520a-3p mimics into CC cells (Fig. 3D and E; *P<0.05, **P<0.01). Furthermore, a luciferase reporter assay was performed to identify the binding of CXCL8 with miR-302c-3p or miR-520a-3p. The results revealed that miR-302c-3p mimics and miR-520a-3p mimics inhibited the luciferase reporter activity of the wild-type CXCL8 3′-UTR construct (Fig. 3F; **P<0.01). Collectively, these data indicated that CXCL8 is a direct common target of miR-302c-3p and miR-520a-3p.

As previously described, CXCL8 has been reported to be involved in cell proliferation-promoting and apoptosis-suppressive behaviour in cancerous cells (23,24). To investigate whether CXCL8 is a tumour-related cytokine in HeLa-S3 and C-33A cells, the effect of CXCL8 on CC cell proliferation was detected by the CCK-8 assay. Briefly, the HeLa-S3 and C-33A cell lines were treated with CXCL8 siRNA or negative control (NC) or with pcDNA3.1-CXCL8 or pcDNA3.1, and then both CCK-8 and EdU staining assays were performed. Compared with the levels in cells treated with si-NC, the mRNA and protein levels of CXCL8 were decreased in cells transfected with si-CXCL8, whereas cells treated with pcDNA3.1-CXCL8 (pcDNA3.1 used as control) showed increased expression (Fig. 4A; **P<0.01). The up- and downregulation of CXCL8 correspondingly increased or decreased the proliferation ability of HeLa-S3 and C-33A cells compared with the controls (Figs. 4B and C and S3A; *P<0.05, ****P<0.0001). To investigate whether CXCL8 can decrease CC cell apoptosis, flow cytometry was also performed. A significant decrease in apoptosis was observed in both HeLa-S3 and C-33A cells transfected with pcDNA3.1-CXCL8, whereas the cells treated with si-CXCL8 showed the opposite effects (Figs. 4D and S3B; ***P<0.001, ****P<0.0001).

To confirm whether CXCL8 downregulation is the mechanism by which miR-302c-3p and miR-520a-3p suppress proliferation and promote apoptosis in CC cells, HeLa-S3 and C-33A cells were co-transfected with pcDNA3.1-CXCL8 or pcDNA3.1 and miR-302c-3p/miR-520a-3p mimics. As expected, restoration of CXCL8 partially blocked the miR-302c-3p and miR-520a-3p-mediated suppression of CC cell proliferation and promotion of apoptosis (Figs. 5A and B and 6; **P<0.01, ***P<0.001, ****P<0.0001, *P<0.05). In summary, these results indicated that CXCL8 is a functional target of miR-302c-3p and miR-520a-3p to suppress the proliferation of CC cells.