The human normal skin keratinocyte line HaCaT and the human cervical cancer cell lines CaSki and SiHa were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA) and were then cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) (both from Gibco, Rockville, MD, USA) and 1% penicillin/streptomycin (Sigma, St. Louis, MO, USA) in accordance with the recommended protocols. The cells were maintained at 37°C in a humidified incubator containing 5% CO₂. Twenty-two clinical specimens of cervical cancer were collected from the Women and Infants Hospital of Zhengzhou. Frozen cancer tissues in liquid nitrogen were stored at -80°C and prepared for RNA isolation. The use of human clinical samples was approved by the Institutional Human Experiment and Ethics Committee of the Women and Infants Hospital of Zhengzhou.

The level of mRNA expression was determined by qRT-PCR analysis. In brief, total RNA from cells or clinical specimens was isolated using the miRNeasy Mini kit (Qiagen, Shanghai, China). The total RNA was reverse-transcribed into cDNA by using M-MLV reverse transcriptase (Takara, Dalian, China) or miScript Reverse Transcription kit (Qiagen). qRT-PCR analysis was performed using SYBR Premix Ex Taq™ II (Takara) with the following primers: FOXQ1 forward, 5′-CGCGGACTTTGCACTTTGAA-3′ and reverse, 5′-AGCTTTAAGGCACGTTTGATGGAG-3′; E-cadherin forward, 5′-TGCCCAGAAAATGAAAAAGG-3′ and reverse, 5′-GTGTATGTGGCAATGCGTTC-3′; vimentin forward, 5′-GAGAACTTTGCCGTTGAAGC-3′ and reverse, 5′-TCCAGCAGCTTCCTGTAGGT-3′; GAPDH forward, 5′-AATGGGCAGCCGTTAGGAAA-3′ and reverse, 5′-TGAAGGGGTCATTGATGGCA-3′; miR-506 forward, 5′-GGGTATTGAGGAAGGTGTT-3′ and reverse, 5′-CAGTGCGTGTCGTGGAGT-3′ and U6 forward, 5′-CTCGCTTCGGCAGCACATATACT-3′ and reverse, 5′-ACGCTTCACGAATTTGCGTGTC-3′. Gene quantification data were normalized to GAPDH for mRNA expression analysis or normalized to U6 for miRNA expression analysis using the 2^−ΔΔCt method.

Proteins were extracted from cells using RIPA lysis buffer containing protease inhibitor cocktail (Beyotime, Haimen, China), and the concentration was measured and quantified using the Bradford method. A total of 25 µg proteins from each cell group were separated via 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The proteins on the gel were transferred onto a nitrocellulose membrane (Millipore, Boston, MA, USA), which was then blocked by blocking solution (3% nonfat-dried milk) for 1 h at 37°C. After blocking, the membrane was incubated with primary antibodies, including anti-FOXQ1 (1:400) and anti-GAPDH (1:10,000) (both from Abcam, Cambridge, UK), which were diluted in blocking buffer at 4°C overnight. The membrane was incubated with horseradish peroxidase-conjugated secondary antibodies (1:2,000; Beyotime) for 1 h at 37°C. The protein bands were detected using chemiluminescence (Amersham, Little Chalfont, UK). Densitometric analysis of the protein bands was performed using Image-Pro Plus 6.0 software (Media Cybernetics, Inc., Rockville, MD, USA).

The expression of FOXQ1 was suppressed by FOXQ siRNA or miR-506 mimic transfection. FOXQ1 siRNA (5′-CGCGGACUUUGCACUUUGA-3′) and negative control siRNA (NC-siRNA; 5′-UUCUCCGAACGUGUCACGU-3′), miR-506 mimics (5′-UAAGGCACCCUUCUGAGUAGA-3′) and negative control miRNA mimics (miR-NC; 5′-UUCUCCGAACGUGUCACGUTT-3′) were synthesized by Shanghai GenePharma Co., Ltd. (Shanghai, China). The miR-506 mimics were transfected into cells at final concentrations of 50 nM using Lipofectamine 2000 (Invitrogen) in accordance with the manufacturer's instructions. For knockdown of FOXQ1, ~1 µg of FOXQ1 siRNA was diluted into 100 µl of DMEM containing 6 µl of Lipofectamine 2000 and incubated for 45 min at room temperature. The siRNA transfection reagent mixture was added to each well and incubated for 7 h. Then, the medium was replaced with fresh medium and cultured for 48 h. The open reading frame of FOXQ1 without 3′-UTR was inserted into pcDNA3.0 plasmids. For the rescue experiment, miR-506 mimics (50 nm) and pcDNA3/FOXQ1 overexpression vectors (3 µg) were co-transfected into cervical cancer cells for 48 h. Empty vectors were used as control.

Cells were grown into 96-well plates (1×10⁴ cells/well), transfected with FOXQ1 siRNA or miR-506 mimics, and then incubated for 24, 48, and 72 h. Cell proliferation was measured using the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide (MTT) method. In brief, 20 µl of MTT solution (5 mg/ml; Sigma) was added to each well after replacement with fresh medium. After 4 h, the formazan crystals were melted by adding 200 µl of dimethyl sulfoxide (DMSO) to each well. After the crystals were dissolved, the optical density of each well was measured at 490 nm using an ELISA reader.

The cDNA fragments of FOXQ1 3′-UTR containing the predicted binding site for miR-506 were inserted into pmirGLO luciferase promoter vectors (Promega, Madison, WI, USA). The cells were seeded into 24-well plates (5×10⁴ cells/well) and were then transfected with 500 ng of pmirGLO-FOXQ1 3′-UTR and 50 nM of miR-506 mimics for 48 h using Lipofectamine 2000. The cells were lysed, and the activity of firefly or Renilla luciferase was detected using a dual-luciferase reporter assay kit (Promega) in accordance with the manufacturer's instructions.

Data are shown as means ± standard deviation. SPSS 11.5 for Windows (SPSS Inc., Chicago, IL, USA) was used for the statistical analyses. Differences between two groups were analyzed using the Student's t-test. One-way ANOVA was conducted in multiple groups for statistical analyses. The correlation between FOXQ1 and miR-506 was evaluated using Spearman's correlation analysis. Statistical significance was considered at P<0.05.

To explore the potential role of FOXQ1 in cervical cancer, we detected FOXQ1 expression in the cervical cancer cell lines CaSki and SiHa through qRT-PCR and Western blot analyses. qRT-PCR data showed that the mRNA expression of FOXQ1 was significantly higher in the CaSki and SiHa cells than that in the HaCaT cells (Fig. 1A). Western blot analysis also revealed a high protein expression level of FOXQ1 in the CaSki and SiHa cells (Fig. 1B). The data indicated that FOXQ1 may function as an oncogene in cervical cancer.

To understand the biological role of FOXQ1 in cervical cancer, we performed loss-of-function experiments in the CaSki and SiHa cells by using siRNA targeting FOXQ1. Cells were transfected with FOXQ1 siRNA, and the expression of FOXQ1 was detected by qRT-PCR and western blot analyses. The mRNA and protein expression levels of FOXQ1 were significantly decreased by FOXQ1 siRNA transfection (Fig. 2A and B). Furthermore, the effect of FOXQ1 silencing on cell proliferation was detected by MTT assay. Transfection of FOXQ1 siRNA markedly impeded the proliferation of CaSki (Fig. 2C) and SiHa (Fig. 2D) cells. These results imply that FOXQ1 is a potential molecular target for inhibiting the proliferation of cervical cancer cells.

To further investigate the function of FOXQ1, we determined the effect of FOXQ1 silencing on the EMT of cervical cancer cells. Results showed that FOXQ1 knockdown significantly increased the expression of the epithelial marker E-cadherin and decreased the expression of the mesenchymal marker vimentin in the CaSki (Fig. 3A) and SiHa (Fig. 3B) cells transfected with FOXQ1 siRNA. The data suggest that FOXQ1 may be a potential target for inhibiting cervical cancer metastasis.

Previous studies have reported that FOXQ1 expression can be regulated by specific miRNAs in cancers (26,27). Bioinformatic research revealed that FOXQ1 has a putative binding site for miR-506 (Fig. 4A), a well-recognized tumor-suppressor miRNA (28,29). A dual-luciferase reporter assay was performed to verify whether or not FOXQ1 is a target gene of miR-506. The results of the luciferase assay showed that miR-506 overexpression significantly inhibited the luciferase activity of the reporter gene with the wild-type FOXQ1 3′-UTR but not with the mutant FOXQ1 3′-UTR in the CaSki (Fig. 4B) and SiHa cells (Fig. 4C). These results indicate that FOXQ1 is a direct target gene of miR-506.

To further confirm that FOXQ1 is the target gene of miR-506, we detected the expression change of FOXQ1 in response to miR-506 mimic transfection. The overexpression of miR-506 through the transfection of miR-506 mimics significantly suppressed the mRNA expression of FOXQ1 in the cervical cancer cell lines (Fig. 5A). Furthermore, the protein expression of FOXQ1 was also significantly decreased by miR-506 overexpression (Fig. 5B).

To further investigate the relationship between miR-506 and FOXQ1, we examined the expression levels of miR-506 and FOXQ1 in clinical specimens and analyzed their correlation. The expression of miR-506 (Fig. 6A) was significantly downregulated in the cervical cancer tissues, whereas that of FOXQ1 (Fig. 6B) was significantly upregulated in the cervical cancer tissues when compared with the adjacent non-tumor tissues. Furthermore, correlation analysis revealed a marked inverse correlation between FOXQ1 and miR-506 expression in the cervical cancer tissues (Fig. 6C). Overall, these results indicate that miR-506 negatively regulates the expression of FOXQ1, and the high expression of FOXQ1 in cervical cancer may be caused by the dysregulation of miR-506.

Considering the regulatory function of miR-506 on FOXQ1 expression as described above, we speculated that miR-506 plays an important role in cervical cancer. To test this hypothesis, we detected the effect of miR-506 overexpression on cervical cancer cell proliferation. MTT assay showed that the proliferation of the cervical cancer CaSki (Fig. 7A) and SiHa (Fig. 7B) cells transfected with the miR-506 mimics was significantly decreased. Furthermore, the transfection of miR-506 mimics significantly increased the expression of E-cadherin and decreased the expression of vimentin in the CaSki (Fig. 7C) and SiHa (Fig. 7D). These results revealed that miR-506 inhibited the proliferation and EMT of cervical cancer cells.

To validate whether miR-506 exerts its tumor-suppressive function by suppressing FOXQ1, we performed rescue experiments. Results showed that the restoration of FOXQ1 expression (Fig. 8A) by transfecting pcDNA3/FOXQ1 overexpression vectors partially rescued the inhibitory effect of miR-506 overexpression on the proliferation (Fig. 8B) and EMT (Fig. 8C and D) in CaSki cells. Similar results were obtained in the SiHa cells (data not shown). Thus, these results revealed that miR-506 functions through FOXQ1 in cervical cancer.