Samples from patients with CC (n=54) admitted to the Department of Gynecology at the Second Affiliated Hospital of Xi'an Jiaotong University (Xi'an) from December 2016-November 2017 were used in the present study. Biopsy samples were obtained by colposcopy prior to surgery, chemotherapy or radiotherapy. Normal cervix (NC) samples were collected from 30 hospitalized age-matched patients with uterine fibroids during hysterectomy. LSIL and HSIL samples were collected from 20 and 24 patients, respectively, undergoing colposcopy and cervical biopsies or after cervical loop electrosurgical excision procedure during the same period. All hematoxylin and eosin (H&E) sections of the specimens were reviewed and confirmed by two pathologists.

The design and implementation of the study were approved by the Ethics Committee of Medical School of Xi'an Jiaotong University (Xi'an, China; no. 2017-266). During the study, a written informed consent was obtained from all participants.

The human CC cell lines HeLa, SiHa, C33A and CaSki were purchased from American Type Culture Collection (Manassas, VA, USA) and Tera-1, which served as a positive control, was a gift from Dr Yue Li (Xi'an Jiaotong University Health Science Center, Xi'an, China). As SOX11 is a stem cell-associated gene, Tera-1 is a teratoma cell line that abundantly expresses stem cell-associated genes. Tera-1 was selected as the positive control. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), except for the CaSki cells which were maintained in McCoy's 5A medium (Sigma-Aldrich; Merck KGaA) supplemented with 10% fetal bovine serum (FBS; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) in a cell culture incubator at 37°C with 5% CO₂.

Cells were trypsinized, counted and seeded in triplicate in 6-wells plate. After 24 h, the medium was replaced with fresh medium containing 0 (PBS), 5, or 10 µM 5-Aza-dC (Sigma-Aldrich; Merck KGaA). The medium was replaced every 24 h. After 72 h, the total cellular RNA and protein were extracted for subsequent experiments.

Genomic DNA was extracted using the Universal Genomic DNA Extraction kit ver. 3.0 (cat. no. DV811A; Takara Biotechnology Co., Ltd., Dalian, China). Genomic DNA (2 µg/sample) was bisulfite-modified, and the bisulfite-modified DNA was purified according to the manufacturer's protocol (EpiTect BisuLfite kit; Qiagen GmbH, Hilden, Germany). Then, the purified bisulfite-modified DNA was amplified using the following primers: SOX11 forward, 5′-AGAGAGATTTTAATTTTTTGTAGAAGGA-3′ and reverse, 5′-CCCCCTTCCAAACTACACAC-3′. The modified DNA was amplified by PCR using the following conditions: 95°C for 3 min, and then 35 cycles of 95°C for 30 sec, 54°C for 30 sec and 72°C for 30 sec, followed by a 10-min incubation at 72°C. The PCR products were analyzed by gel electrophoresis in 2.5% agarose to confirm that a single band had been produced. Then, TA clones were established according to the instructions of the protocol for the pEASY-T1 Cloning kit (Beijing Transgen Biotech Co., Ltd., Beijing, China). The 10–15 positive monoclonal bacteria solution were sequenced by Xi'an Qing Biological Co., Ltd. (Xi'an, China). The sequencing primers were M13 forward, 5′-GTTTTCCCAGTCACGAC-3′ and reverse, 5′-CAGGAAACAGCTATGAC-3′. The sequencing results were analyzed using the BiQ Analyzer 2.0 (Max Planck Institute for Informatics, Saarbrücken, Germany) and were output as circle graphs.

Total RNA was extracted using the RLT reagent with 1% β-mercaptoethanol in the RNeasy Mini Kit (cat. no. 74106; Qiagen GmbH) according to the manufacturer's protocol. Total RNA (500 ng) was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (cat. no. 4368814; Applied Biosystems; Thermo Fisher Scientific, Inc.) with incubation for 10 min at 25°C, 120 min at 37°C and 5 min at 85°C. qPCR was performed on an Applied Biosystems 7700 Prism RT-PCR machine (Applied Biosystems; Thermo Fisher Scientific, Inc.) and conditions were as follows: Enzyme activation for 10 min at 95°C, PCR cycle denaturation for 15 sec at 95°C and annealing/elongation 1 min at 60°C. The sequences of the SOX11 primers were as follows: Forward, 5′-GGTGGATAAGGATTTGGATTCG-3′ and reverse, 5′-GCTCCGGCGTGCAGTAGT-3′. Expression was normalized to the expression of 18S and transformed using the relative standard curve method and comparative quantification cycle (Ct) method (ΔΔCq) as described previously (27).

Cells and CC cell line samples were washed twice with cold PBS, followed by lysis buffer (cat. no. P0013D; Beyotime Institute of Biotechnology, Haimen, China). Protein concentration was quantified using a bicinchoninic acid kit (cat. no. P0009; Beyotime Institute of Biotechnology) according to the manufacturer's instructions. Protein lysates (25 µg) were separated at 80 V on 10% acrylamide gel for ~2 h. Transfer to Immobilon-FL (EMD Millipore, Billerica, MA, USA) membrane was performed at 20 V for 1.5 h. Following blocking in the Odyssey blocking buffer (cat. no. 927-40000; LI-COR Biosciences, Lincoln, NE, USA) for 50 min at room temperature, a primary antibody rabbit polyclonal anti-human SOX11 (1:500 dilution; cat. no. sc-20096; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), GAPDH (1:10,000 dilution; cat. no. G8795; Sigma-Aldrich; Merck KGaA) and β-tubulin (1:5,000 dilution; cat. no. sc-80011; Santa Cruz Biotechnology) were incubated at 4°C overnight. Secondary antibody conjugated to Alexa Fluor^® 680 dye (cat. no. A32734; Invitrogen; Thermo Fisher Scientific, Inc.) or IRdye800 (cat. no. 610-731-002; Rockland Immunochemicals Inc., Limerick, Pennsylvania, USA) was subsequently incubated with the membrane for 1 h at room temperature to visualize the proteins at 700 or 800 nm using a LI-COR Odyssey imaging system (LI-COR Biosciences). The results were quantitatively analyzed using the ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Human specimens were fixed with 4% paraformaldehyde overnight at room temperature, then embedded in paraffin and sectioned into 4 µm slices. The slides were baked at 65°C overnight prior to deparaffinization in xylene twice for 20 min and hydrated in a series of graded ethanol (100, 95, 90 and 80% ethanol; 5 min each). The sections were boiled in citrate buffer (pH 6.0) by heating in a pressure cooker for 1–2 Alexa Fluor^® 680 min for antigen retrieval. Endogenous peroxidases were blocked for 10 min with freshly prepared 3% H₂O₂ at room temperature. Following blocking with goat serum for 30 min at room temperature, the SOX11 primary antibody (1:200 diluted in 1% bovine serum albumin PBS solution; cat. no. sc-20096; Santa Cruz Biotechnology, Inc.) was incubated with the slides at 4°C overnight. The following day, the slides were incubated with a secondary antibody labeled with horseradish peroxidase (1:50 dilution, cat. no. GAR-HRP; Pierce; Thermo Fisher Scientific, Inc.) for 40 min at room temperature. Diaminobenzidine was used to color the slides for ~20 min at room temperature. The sections were observed under the microscope to control the reaction time when the sections were incubated with the DAB reagent for signal amplification. Then the sections were stained with H&E. Histological analysis was performed by two blinded pathologists. At least 10 high-power fields at ×1,000 magnification were examined for each sample. The staining score was classified into four grades based on the staining intensity as follows: 1, no cell staining; 2, weak yellow; 3, moderate yellow; and 4, strong brown yellow. The proportion of positively stained cells was classified into four levels as follows: 1 (0–25%), 2 (26–50%), 3 (51–75%) and 4 (76–100%). The immunoreactivity score (IRS) was obtained by multiplying the intensity and proportion values, and samples were grouped into three levels as follows: Negative (−, score 1–4), positive (+, score 5–9), strongly positive (++, score 10–16). The mean score was used for comparison between groups.

Cell viability was measured using the PrestoBlue kit (Invitrogen; Thermo Fisher Scientific, Inc.). A total of 2,000 cells in the logarithmic growth phase were seeded into a 96-well plate in triplicate and allowed to adhere overnight. Throughout the 6-day period, the medium was replaced with fresh medium every day, and the same concentration of 5-Aza-dC was added. PrestoBlue reagent was used to assess the proliferative ability according to the standard manufacturer's protocol. The fluorescence was read at a wavelength of 570 nm using a FLUOstar Optima microplate spectrophotometer (BMG Labtech GmbH, Ortenberg, Germany).

Statistical analyses were performed using the Prism 7 software (GraphPad Software Inc., La Jolla, CA, USA) and SPSS version 22.0 for Windows (IBM Corp., Armonk, NY, USA). The unpaired t-test with Welch's correction was used to analyze the difference between two groups. The two-tailed Mann Whitney U test with Bonferroni's correction for post hoc comparisons was performed following the Kruskal-Wallis test and Tukey's post hoc test was used following the one-way analysis of variance test to evaluate the statistical comparison of multiple groups. The Pearson χ² analysis was used to analyze the association between SOX11 expression and the clinical features, when the number was <5, Fisher's exact test was used. Pearson's correlation test was used to analyze the correlation between the SOX11 mRNA expression level and its promoter methylation level. Error bars represent ± standard error. P<0.05 was considered to indicate a statistically significant difference.

SOX11 has been demonstrated to have different properties in different human cancers. It is upregulated in gastric (28) and prostate cancer (29), and other types of cancers, whereas it is downregulated in medulloblastoma (30) and malignant gliomas (20). To determine the expression of SOX11 during the development of CC, SOX11 protein expression level was analyzed by immunohistochemistry and western blot analysis in NC, LSIL, HSIL and CC tissues. SOX11 was localized in the nuclei of all positive cells with different levels in different specimens. SOX11 was highly expressed in NC and LSIL epithelial basal cells, and weakly expressed or virtually absent in tumor parenchymal cells of HSIL and CC. The IRS of SOX11 was as follows: NC, 9.933±0.948; LSIL, 7.200±0.766; HSIL, 4.917±0.492; and CC, 3.074±0.301 (Fig. 1B). With the progression of cervical lesions, the expression level of SOX11 gradually decreased, and there were statistical differences in IRS levels of SOX11 in the four groups (NC vs. LSIL, P=1.000; NC vs. HSIL, P=0.022; NC vs. CC, P<0.0001; LSIL vs. HSIL, P=0.625; LSIL vs. CC, P<0.0001; HSIL vs. CC, P=0.057; P-value are adjusted). In addition, the expression level of SOX11 was determined by western blot analysis in randomly selected tissues including 8 NC and 20 CC (Fig. 1C). The relative expression of the SOX11protein is presented as mean SOX11 expression of 0.95±0.07 in NC, and 0.17±0.04 in CC (Fig. 1D). Thus, the expression of SOX11 protein in NC was 5.59 fold higher than that in CC specimens (P<0.001). The association between SOX11 expression based on the immunohistochemistry staining results and the clinicopathological characteristics in the patients with CC are summarized in Table I. The Pearson χ² analysis revealed a significant association between SOX11 expression and the tumor grade (P=0.005) and HPV status (P<0.001); however, there was no significant association between SOX11 expression and other clinical features, including age, International Federation of Gynecology and Obstetrics stage, lymph node, myometrial invasion depth and histologic type. The association between SOX11 expression and the HPV status in the patients with LSIL and HSIL in Tables II and III reveal a significant association between SOX11 expression and HPV status (P=0.032 and 0.035, respectively). These findings suggest that the expression of SOX11 decreases with the development of CC malignancy, which strongly suggested that the loss of SOX11 function, which may act as a tumor suppressor, promotes the progression of CC.

The hypermethylation of promoter CpG islands is one of the essential mechanisms of transcriptional silencing of TSGs, and also the most important early, common event in the process of cervical disease progression to cancer (13,31–34). To further investigate the mechanism of the downregulation of SOX11 during the development of CC, 10 NC and 10 CC specimens were randomly selected to determine the methylation status of the SOX11 promoter. Four CpG islands of SOX11 were identified with a GC content >50% and observed/expected CpG ratio >0.6 when to promoter region (2,000 bp upstream of the transcription start site) of SOX11 was analyzed using MethPrimer 2.0 (urogene.org/methprimer2/; Fig. 2A). Subsequent sequencing experiments were performed on the fourth CpG island, which had previously been reported to be determinative for SOX11 expression in multiple cell lines (25) and includes 28 CpG sites (Fig. 2A). The results are presented in Fig. 2B, with the CpG sites are indicated with circles (solid circle indicate methylation and unshaded circle indicates no methylation). The SOX11 promoter was hypermethylated in CC (total mean level was 85.71%), which was significantly higher than that in the NC samples (total mean level was 12.68%) at each CpG site (Fig. 2C). The mRNA expression level of SOX11 in CC was significantly lower than that in NC tissues (P<0.001; Fig. 2D). Pearson's correlation analysis of the SOX11 mRNA expression level and its promoter methylation level revealed that the SOX11 mRNA expression level in CC was negatively correlated with hypermethylation in the promoter region (r=−0.8080; P<0.001; Fig. 2E). The expression of SOX11 protein was downregulated, with its promoter region hypermethylated in CC. These results suggest that hypermethylation of the SOX11 promoter may be involved in the mechanism of downregulation of SOX11 in CC.

The transcriptional and translational level of SOX11 and its promoter methylation status was examined in CC cell lines. A marked difference was observed between the SOX11 expression and promoter methylation level. The Tera-1 cell line as a positive control. The results revealed high levels of SOX11 promoter methylation in HeLa, C33A, CaSki and SiHa (90.71, 77.14, 99.64 and 98.21%, respectively) cell lines, consistent with a lack of SOX11 mRNA and protein expression. By contrast, SOX11 promoter methylation was low in the Tera-1 (10.36%), and its mRNA and protein expression was significantly higher than those in the four CC cell lines (Fig. 3). These findings suggest that SOX11 expression in the CC cell lines is negatively associated with the methylation level of the promoter, which is consistent with the results in human tissue specimens.

As the silencing of TSGs by aberrant methylation of the gene promoter is reversible (35), to further investigate the role of hypermethylated promoter level in the expression regulation of SOX11, HeLa, C33A, CaSki and SiHa cell lines were treated with different doses of the demethylating agent 5-Aza-dC. The mRNA and protein expression levels of SOX11 under different conditions were measured by RT-qPCR and western blot analysis. As shown in Fig. 4A, when DNA of cervical cells was demethylated with different doses of 5-Aza-dC, the SOX11 mRNA level was increased from 0.95±0.05 to 15.65±0.65 in HeLa, 0.95±0.07 to 34.30±0.70 in C33A, 1.12±0.11 to 57.85±1.05 in CaSki and 1.00±0.12 to 31.2±0.80 in SiHa cells treated with 10 µM compared with no 5-Aza-dC treatment. Similarly, the SOX11 protein expression level was gradually increased from 0.78 to 1.42 in HeLa, 0.52 to 1.28 in C33A, 0.38 to 1.36 in CaSki and 0.70 to 1.38 in SiHa cells comparing no 5-Aza-dC treatment to 10 µM. Cell viability was also to determine the contribution of the SOX11 expression level to CC cell growth. The proliferation ability of cervical cells was significantly suppressed by 5-Aza-dC (Fig. 4B). These results suggest that the hypermethylation of the promoter reduces SOX11 expression in the four CC cell lines and activation of SOX11 by demethylation of the promoter significantly inhibited the proliferation of cervical cells.