The cervical cancer cell lines CaSki and HeLa were purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in RPMI 1640 medium (Cambrex, Walkersville, MD, USA) supplemented with 10% fetal bovine serum (FBS; HyClone; GE Healthcare Life Science, Logan, UT) at 37°C and 5% CO₂. Transfection of cells with miR-326 was carried out using Lipofectamine™ 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), according to the manufacturer's protocol.

A total of 25 primary cervical cancer specimens and 25 non-cancerous specimens were collected from female patients who had undergone surgical treatment at The Second Hospital of Shandong University (Jinan, China) between January 2014 and September 2014. The mean of the patients was 47 years (range, 40–55 years). The samples were processed (liquid nitrogen frozen) and stored in RNAlater (Qiagen, Inc., Valencia, CA, USA) at −20°C prior to extraction of RNA. The present study was approved by the Bioethics Committee of Shandong University (Jinan, China), and written informed consent and approval were provided by each patient prior to the study.

RNA was isolated from tissue and cell samples using TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. A polyA tail was added to the mature miRNA template, and complementary DNA (cDNA) was synthesized using a polyT primer with a 3′degenerate anchor and a 5′ universal tag. RT-qPCR was carried out using a miRCURY LNA™ Universal RT microRNA PCR system with SYBR Green master mix and LNA-based primer sets for target miRNAs (Homo sapiens-miR-326; cat. no. 204512) and U6 small nuclear RNA (cat. no. 203907; all from Exiqon, Inc., Woburn, MA, USA). PCR cycles were as follows: 94°C for 3 min, followed by 40 cycles of 94°C for 30 sec, 60°C for 30 sec and 72°C for 30 sec. The relative levels of miRNAs in the cells and tissues were normalized to U6 and were calculated using the 2^−ΔΔCq method (25). To quantify the level of ELK1 mRNA expression, 1 µg total RNA was reverse transcribed to cDNA using PrimeScript RT reagent kit (Takara Biotechnology Co., Ltd., Dalian, China), and qPCR was carried out using the reverse-transcribed product, SYBR Green dye mix (Invitrogen; Thermo Fisher Scientific, Inc.) and specific primers for ELK1 and GAPDH, which were as follows: ELK1, 5′-CCTTCTATCAGCGTGGAT-3′ (forward) and 5′-GTGGTGGTGGTAGTAGTC-3′ (reverse); and GAPDH, 5′-ACCCAGAAGACTGTGGATGG-3′ (forward) and 5′-CAGTGAGCTTCCCGTTCAG-3′ (reverse). The thermo cycling conditions for the RT-qPCR were as follows: 95°C 10 min, 40 cycles of 95°C for 20 sec, 55°C for 30 sec, 72°C for 20 sec. The relative level was calculated using the 2^−ΔΔCq and the level of ELK1 mRNA was normalized to that of GAPDH.

Proteins were extracted using RIPA lysis buffer (Beyotime Institute of Biotechnology, Shanghai, China) and the concentration was measured using the Bio-Rad protein assay kit (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) according to the manufacture's protocol. A total of 30 µg protein was separated using 12.5% SDS-PAGE, and then transferred onto polyvinylidene difluoride membranes. The membranes were blocked with 5% non-fat milk at room temperature for 1 h and incubated with primary antibodies against ELK1 (cat. no. 14507-1-AP; dilution, 1:1,000; Proteintech) and GAPDH (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) at 4°C overnight. The membranes were then incubated with horseradish peroxidase-conjugated secondary antibodies (cat. no. ab6721; dilution, 1:2,000; Abcam). The blots were examined using an enhanced chemiluminescence detection system according to the manufacture's protocol (Pierce; Thermo Fisher Scientific, Inc.). The band intensities of the western blotting were analyzed using Image Analysis Software v2.0 (Thermo Fisher Scientific, Inc.).

MTT and 5-ethynyl-2′-deoxyuridine (EdU) incorporation assays were used to determine cell proliferation. Cells were seeded into 96-well plates at 4×10⁴ cells/well in triplicate for each transfection group: miR-326 mimic, miR-326 inhibitor and blank control. MTT was added in each well (5 mg/ml) at various time points (24, 48 and 72 h) and then incubated in the dark at 37°C for 2 h. Absorbance was determined at a wavelength of 570 nm. An EdU incorporation assay was performed to analyze cell proliferation using a Cell-Light™ EdU Apollo^® 488 In Vitro Imaging kit (Guangzhou RiboBio Co., Ltd., Guangzhou, China), according to the manufacturer's protocol.

Cells (5×10⁵ cells/ml) were seeded in 6-well plates and cultured in complete medium (RPMI-1640 medium with 10% FBS). Following 24 h of culture, when the cells had reached 90–100% confluence, a single wound was created in the center of the well by removing the attached cells with a sterile 200-µl pipette tip. After 56 h of culture, cells that had migrated into the wounded area were visualized and images were captured under an inverted microscope. Each experiment was carried out at least three times independently. Wound closure was monitored over time and the percentage of closure was determined.

A Matrigel invasion chamber was used to determine cell invasive capacity. A cell suspension (5×10⁵ cells/ml) was prepared in serum-free medium (RPMI-1640 medium with 0.1% FBS) for the transfected CaSki cells (transfected with miR-326 mimic, inhibitor or NC). A 300-µl volume of cell suspension was added into the upper chamber coated with Matrigel, and 500 µl of Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS was added into the lower chamber. The cells were then incubated at 37°C with 5% CO₂ for 24 h. Cells on the upper surface were removed, whereas those on the lower surface were stained by 1% eosin for 30 min. The cell number was determined in at least five randomly selected fields under a microscope (SZ61; Olympus, Tokyo, Japan).

The following online miRNA target prediction algorithms were used to evaluate the potential target genes of miR-326: PicTar (http://www.pictar.org/), TargetScan (http://www.targetscan.org/vert_71/), Microcosm Targets (http://www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5/) and miRanda (http://www.microrna.org/microrna/home.do).

CaSki cells in a 96-well plate were transfected with 50 nM miR-326 mimic or negative control miRNA and then co-transfected with the wild-type or mutant 3′-UTR of the ELK1 gene with 0.2 mg/ml vector (Ambion; Thermo Fisher Scientific, Inc.). Following 48 h of transfection, luciferase activity was determined using the Dual-Luciferase^® Reporter Assay System (Promega Corporation, Madison, WI, USA). Firefly luciferase activity was then normalized to the corresponding Renilla luciferase activity. Luciferase assays were repeated in three independent experiments.

In order to construct ELK1 overexpressing plasmid, the gene without 3′-UTR was amplified from cDNA of human normal cervical tissues by polymerase chain reaction. The sequences of primers for cloning were: F, CTAGCTAGCATGGACCCATCTGTGACGCTGTG; R, CCGCTCGAGTCATGGCTTCTGGGGCCCT. The PCR products were cloned into the pMD18-T vector (TAKARA Biotechnology Co., Ltd.). Then, pCMV6-ELK1 was obtained by subcloning a NheI-XhoI fragment from pMD18-T/ELK1 into the NheI-XhoI site of pCMV6-Entry Vector.

Results are presented as the mean ± standard deviation of three independent experiments. A Student's t test was used to compare the experimental groups with the controls, whereas two-way analysis of variance followed with Bonferroni's multiple comparison test was used to determine the differences among three or more experimental groups. P<0.05 was considered to indicate a statistically significant difference. SPSS 18.0 software (SPSS Inc., Chicago, IL, USA) was used to analyze the data.

miR-326 expression was determined in 25 fresh cervical cancer tissue specimens and 25 healthy control specimens. The results of RT-qPCR demonstrated that miR-326 expression was downregulated in cervical cancer tissues compared with that in non-neoplastic samples (P=0.0109; Fig. 1). Significantly decreased miR-326 expression was also demonstrated in cervical cancer cell lines (HeLa, P=0.0316; and CaSki, P=0.0045; Fig. 1) using RT-qPCR. In spite of the variation in the degree of downregulation of miR-326 being observed in the cell lines, the expression was relatively low. As downregulation was more evident in the CaSki cell line than in the HeLa cell line, CaSki cells were used for subsequent experiments.

As miR-326 expression was significantly downregulated in cervical cancer tissues, it was hypothesized that it may be involved in tumor cell proliferation. Therefore, the effect of the overexpression or low expression of miR-326 on the proliferation of cervical cancer cell lines (CaSki and HeLa) was investigated. The results of an MTT assay revealed that transfection with miR-326 mimic inhibited the proliferation of cervical cancer cells (Fig. 2A, Caski: P_24h=0.0416; P_48h=0.0356; P_72h=0.0528; Hela: P_24h=0.00245; P_48h=0.00986; P_72h=0.0238). Cervical cancer cells transfected with the miR-326 inhibitor demonstrated enhanced cell proliferation compared with that of the control (Fig. 2A, Caski: P_24h=0.0969; P_48h=0.00221; P_72h= 0.0055; Hela: P_24h=0.011; P_48h=0.0104; P_72h=0.0841), which was similar to the results obtained using an EdU incorporation assay (Fig. 2B). These results indicated that miR-326 serves an important role in suppressing cervical cancer proliferation.

Migration and invasion assays were carried out to investigate the effects of miR-326 on cervical cancer cells in vitro. Cervical cancer cells in the miR-326 mimic group exhibited decreased migratory and invasive abilities compared with those displayed by control cells (Fig. 2C). Conversely, the miR-326 inhibitor promoted the migration and invasion of cervical cells (Fig. 2C). These effects were determined to be statistically significant in each cell type (Fig. 2D).

To reveal the molecular mechanism underlying the inhibition of the proliferation and metastasis of cervical cancer cells by miR-326, a combination of four bioinformatics algorithms was used (TargetScan, PicTar, miRanda and Microcosm Targets) to search for accurate potential targets of miR-326. All four approaches predicted ELK1 as a potential target of miR-326, and the 3′-UTR of ELK1 mRNA contained a highly conserved binding site between position 1,261 and position 1,268 of the miR-326 seed sequence (the core sequence that encompasses the first 2–8 bases of the mature miRNA; Fig. 3A). The wild-type and mutant ELK1 3′-UTRs were generated with sequences presented in Fig. 3A. A dual-luciferase reporter assay was performed on CaSki cells, and luciferase activity was significantly decreased in CaSki cells co-transfected with the wild-type ELK1 3′-UTR and miR-326 mimic (P=0.0010; Fig. 3B). This result indicates that miR-326 directly targets the 3′-UTR of ELK1 in CaSki cells.

To investigate the association between miR-326 and ELK1 in human cervical cancer, ELK1 expression was determined in 25 cervical cancer tissues and 25 wild-type cervical tissues using RT-qPCR. ELK1 mRNA levels in cervical cancer tissues were significantly increased (P=0.0125; Fig. 3C) compared with those in wild-type cervical samples. The upregulation of ELK1 was associated with the downregulated expression of miR-326 in 25 cervical cancer tissues (Fig. 1). Therefore, the downregulation of miR-326 may contribute to the overexpression of ELK1 in human cervical cancer.

The role of miR-326 in the regulation of ELK1 mRNA expression levels was further investigated in the CaSki cervical cancer cell line. Following transfection of CaSki cells with scrambled miRNA (negative control), miR-326 mimic or miR-326 inhibitor, the expression level of miR-326 in each group was determined to confirm that the transfection efficiency was satisfactory (Fig. 4A). Following transfection with the miR-326 mimic, the expression level of miR-326 was significantly upregulated (P=0.0005; Fig. 4A) compared with that in the control group. Following transfection with the miR-326 inhibitor, the expression level of miR-326 was significantly decreased (P=0.0190; Fig. 4A) compared with that in the control group. The mRNA and protein levels of ELK1 in each group were determined using RT-qPCR and western blot analysis, respectively. Upregulation of miR-326 induced a decrease in the mRNA (Fig. 4B) and protein (Fig. 4C) expression levels of ELK1, whereas the downregulation of miR-326 resulted in the upregulation of the mRNA (Fig. 4B) and protein (Fig. 4C) expression levels of ELK1 in CaSki cells. Therefore, the expression levels of ELK1 appear to be negatively regulated by miR-326 in CaSki cells.

To determine whether ELK1 is the real functional target of miR-326, ELK1 expression plasmids were constructed without the 3′-UTR of ELK1 (pCMV6-ELK1), which are not able to be regulated by miR-326, to perform a rescue experiment. RT-qPCR analysis demonstrated that overexpression of ELK1 is able to restore the ELK1 mRNA level decreased by miR-326 (Fig. 5A). Furthermore, MTT, migration and invasion assays were performed in CaSki and HeLa cells co-transfected with miR-326 mimic plus pCMV6-ELK1 or empty vector, or with negative control plus empty vector. Restoration of ELK1 eliminated the cell viability, invasion and migration that was decreased by miR-326 mimic (Fig. 5B-E). These results demonstrated that ELK1 is a direct functional target gene of miR-326 and that miR-326 functions as a tumor suppressor through ELK1.