miR-25-3p reverses epithelial-mesenchymal transition via targeting Sema4C in cisplatin-resistance cervical cancer cells.

Acquisition of epithelial-mesenchymal transition (EMT) has recently been proposed as an important contributor of drug resistance in cervical cancer cells. However, the underlying mechanisms are still unclear. MicroRNAs play a crucial role in regulating EMT. The aim of this study was to explore the potential role of miR-25-3p in regulating EMT in cisplatin-resistant (CR) cervical cancer cells. To this end, we established stable CR cervical cancer cells, HeLa-CR and CaSki-CR, and investigated the function of miR-25-3p in regulating EMT. It is found that CR cervical cancer cells possessed more EMT characteristics and demonstrated higher migratory abilities and invasiveness. miR-25-3p downregulation was also seen in HeLa-CR and CaSki-CR cells. Of note, ectopic expression of miR-25-3p reversed the EMT phenotype and sensitized CR cells to cisplatin via targeting Sema4C. Furthermore, stable overexpression of miR-25-3p in HeLa-CR cells suppressed tumor growth in mice, downregulated Sema4C and Snail, and upregulated E-cadherin compared with the control group. These results suggest that miR-25-3p is an important regulator of cervical cancer EMT and chemoresistance. Thus, upregulation of miR-25-3p could be a novel approach to treat cervical cancers that are resistant to chemotherapy.

Cervical cancer (CC), a common malignancy in gynecology, is one of the main causes of cancer‐related mortality among women worldwide.1 The prognosis of patients with advanced/recurrent cervical cancer is extremely poor, with the 1‐year survival being only 10–20%.2 Cervical cancer patients are standardly treated with chemotherapy. Cisplatin, a small‐molecule platinum compound, has shown promise in treating advanced/recurrent cervical cancer.3 Cisplatin‐mediated anticancer effect is linked to multiple intertwined signaling pathways.4 However, resistance to cisplatin, which is acquired intrinsically or during cancer progression, may seriously compromise the efficacy of cisplatin.

Emerging bodies of evidence have indicated the essential role of epithelial‐mesenchymal transition (EMT) in the progression of human cancers.5 It is known that during EMT, epithelial cells gain mesenchymal phenotype, resulting in enhanced invasion and metastasis.6 Concomitantly, epithelial cells downregulate epithelial markers such as E‐cadherin, meanwhile acquiring mesenchymal markers including Vimentin, Snail and Slug.7 Growing evidence indicates that EMT is closely associated with drug resistance.8 For example, the transcription factor Twist1, one of the EMT inducers, was demonstrated to be involved in ovarian cancer metastasis and chemoresistance.9 Recent studies suggest the involvement of EMT‐associated transcription factors in chemoresistance in human breast cancer and cervical cancer cells.10, 11

MicroRNAs (miRNAs) are critically involved in the regulating drug resistance and EMT.12 It has been found that upregulation of miR‐200 and let‐7 resulted in the reversal of EMT in gemcitabine‐resistant pancreatic cancer cells.13 The overexpression of miR‐200c and its target, mitogen‐inducible gene 6, are in close correlation with EMT and resistance to erlotinib.14 Additionally, suppression of miR‐137 in a drug‐resistant SCLC cell line increased its sensitivity to cisplatin.15 These findings suggested the important role of miRNAs in regulating chemotherapy‐induced EMT. Recent studies reported that miR‐25‐3p regulates carcinogenesis in a variety of cancers, including breast cancer,16 cholangiocarcinoma17 and ovarian cancer.18 Interestingly, the serum concentration of circulating miR‐25‐3p was also associated with ovarian cancer drug‐resistance.19 However, little is known about whether miR‐25‐3p is involved in regulating chemotherapy and EMT in human cervical cancer. Herein, we explored the role of miR‐25‐3p in regulating cisplatin‐resistant induced EMT in cervical cancer.

Materials and Methods

Cell culture and reagents

Human cervical cancer cell lines CaSki and HeLa were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and cultured in RPMI‐1640 supplemented with 10% heat‐inactivated fetal bovine serum (Hyclone, Logan, UT, USA). HeLa and CaSki cells were cultured in increasing concentrations of cisplatin for over 6 months to establish cisplatin‐resistant (CR) cell lines, HeLa‐CR and CaSki‐CR. All cells were cultured in a humidified atmosphere at 37°C and 5% CO2. Mycoplasma contamination of cell culture was regularly checked every 3 months. Cisplatin, MTT (3‐[4,5‐dimethyl‐2‐yl]‐2,5‐diphenyl tetrazolium bromide) and all other chemicals used in this study were purchased from Sigma (St. Louis, MO) unless stated otherwise.

MTT assay

The cell viability was evaluated using MTT assay. In brief, the cells were seeded into 96‐well plates (1 × 103 cells/well) and cultured overnight. Then, cells were treated with different concentrations of cisplatin for 72 h. At the end of incubation, 50 μL of MTT (5 mg/mL) were added onto cells, followed by dissolving the resulting formazan crystals in 100 μL of dimethylsulfoxide (DMSO). The cell viability was calculated using as follows: cell viability (%) = OD of treated cells/OD of control cells × 100. Cell growth inhibition (%) = 100 − cell viability (%). All assays were performed with at least three independent experiments.

Cell attachment and detachment assay

The cells were seeded in 24‐well plate (5 × 104 cells/well). Briefly, for attachment assay, after 1 h of incubation, non‐attached cells removed by washing the culture plate twice with phosphate‐buffered saline (PBS), and the attached cells were trypsinized and counted. The attachment data were quantified as a percentage of the number of attached cells versus the total cell number. For detachment assay, the cells were detached with 0.05% trypsin (Invitrogen, Carlsbad, CA, USA) for 3 min and counted after 24 h. The remaining attached cells were further trypsinized with 0.25% trypsin and counted. Cell detachment data were expressed as the ratio of the number of detached cells to the total cell number.

Colony formation assay

Cells, at the density of 200 cells/well, were plated in 6‐well culture plates. Two wells of cells were used for each group. After the incubation of 14 days at 37°C, cells were washed twice with PBS and stained with hematoxylin. The number of colonies with more than 50 cells was counted under a light microscope. The colony formation efficiency was represented as (number of colonies/number of cells seeded) × 100%. All assays were independently performed in triplicate.

Wound‐healing assay

Cells (1 × 106 cells per well) were plated on 6‐well plates. Cells were cultured in the serum‐free medium containing 1% bovine serum albumin (BSA). A sterile 200 μL pipette tip was used to enforce a gap by removing cells. The migration process was monitored (after identification of each wounded zone) in six areas, immediately and 48 h after wounds were made, using an inverted microscope (Nikon TMS‐F, 301655, Nikon, Tokyo, Japan) installed with a digital camera (Nikon Digital shot DS‐L1, Nikon). Cell migration data were expressed as the ratio of change in gap width divided by initial gap width.

Cell invasion assays

For the cell invasion assay, 1 × 104 cells in 100 μL serum‐free medium were seeded in a transwell apparatus (Costar, Corning, NY, USA) containing a fibronectin‐coated polycarbonate membrane insert. Medium (500 μL) containing 10% fetal bovine serum was added in the lower chamber as chemoattractant. After the 8‐h incubation at 37°C in a 5% CO2 atmosphere, the insert was washed extensively with PBS and the top surface of the insert were wiped clean with a cotton swab to remove cells. Cells on the lower surface were fixed with methanol, stained with crystal violet. Cells adhering to the lower surface were counted using five predetermined fields (×100). All assays were independently repeated in triplicates.

RNA isolation, reverse transcription and qRT‐PCR

Total RNA was isolated from cells by Trizol Reagent (Invitrogen). For miR‐25‐3p detection, cDNA was synthesized using the PrimeScript RT reagent Kit (TaKaRa, Dalian, China). Quantitative reverse transcription‐polymerase chain reaction (qRT‐PCR) was carried out using SYBR Premix ExTaq (TaKaRa) with the Stratagene Mx3000P real‐time PCR system (Agilent Technologies, Inc., Santa Clara, CA, USA). The endogenous snRNA U6 was used as house‐keeper gene to normalize the expression levels of other genes. The relative expression of miR‐25‐3p was quantified with the 2−ΔΔCT method. For Sema4C mRNA analyses, Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI, USA) was used for cDNA synthesis. The mRNA level of other genes was normalized to the mRNA level of GAPDH, and was calculated by the 2−ΔΔCT method. PCR reactions for each gene were performed in triplicates. The following primers were used in PCR reaction: Sema4C, forward: 5′‐ACC TTG TGC CGC GTA AGA CAG‐3′, reverse: 5′‐CGT CAG CGT CAG TGT CAG GAA‐3′; GAPDH, forward primer: 5′‐CAG CCT CAA GAT CAT CAG CA‐3′, reverse: 5′‐TGT GGT CAT GAG TCC TTC CA‐3′.

Western blot analysis

To extract protein from cells, cells grown on plates were trypsinized and detached cells were collected by centrifugation. Cell pellet was washed with PBS, and lysed with cold lysis buffer supplemented with protease inhibitors. Protein (30 μg) from each cell lysate was equivalently loaded on the precast gel, and used for electrophoresis. Gels were subsequently blotted onto nitrocellulose membranes (0.45 μM; Bio‐Rad , Hercules, CA), followed by blocking of non‐specific binding with a solution containing 1 × PBS, 0.1% Tween‐20, and 5% non‐fat dry milk powder, at room temperature for 1 h. Membranes were incubated with the primary antibodies against anti‐E‐cadherin, anti‐Snail, anti‐Vimentin, anti‐β‐Actin (Cell Signaling Technology, Danvers, MA, USA) and anti‐Sema4C (Abcam, Cambridge, MA, USA) at 4°C overnight. After extensive washing with TBST, horseradish peroxidase (HRP)‐conjugated secondary antibodies (BioRad) were applied to the membrane and incubated for 1 h at room temperature. Bands were detected with the enhanced chemiluminescence kit (SuperSignal West Pico substrate; Pierce, Rockford, IL, USA). Quantification of signal intensities was performed by densitometry on a Xerox scanner using NIH ImageJ software (ImageJ, Bethesda, MD).

Transfection

To evaluate the effect of Sema4C knockdown, cells, which were seeded in 6‐well plates, were subjected to transfection with Sema4C siRNA, or control siRNA using Lipofectamine 2000. The sequences used for Sema4C siRNA are as followed: Sema4C siRNA, forward, 5′‐GGA GCA UGG AGA GUU UGA ATT‐3′, reverse, 5′‐UUC AAA CUC UCC AUG CUC CTT‐3′; non‐coding siRNA, forward, 5′‐UUC UCC GAA CGU GUC ACG UTT‐3′, reverse, 5′‐ACG UGA CAC GUU CGG AGA ATT‐3′.

To evaluate the effect of miR‐25‐3p overexpression, cells were transfected with miR‐25‐3p mimic or the nonspecific control (GenePharma, Shanghai, China) using lipofectamine RNAiMAX reagent (Invitrogen) following the manufacturer's protocol. MiR‐25‐3p mimic: Sense 5′‐UCC CUG AGA CCC UAA CUU GUG A‐3′; antisense 5′‐ACA AGU UAG GGU CUC AGG GAU U‐3′. The cells were then used for further analysis.

The cells seeded in 24‐well plates were used for transfection with miR‐25‐3p inhibitor (for knock‐down of miR‐25‐3p) or the nonspecific control using DharmaFect Transfection Reagent (Dharmacon, Lafayette, CO) following the manufacture's protocol. The cells were subjected to analysis by luciferase reporter assay.

3′‐UTR luciferase reporter assays

For reporter assays, a fragment of Sema4C 3′‐UTR was amplified by PCR and cloned into psiCHECK‐2 vectors (named wt). The miR‐25‐3p binding site within Sema4C 3′‐UTR was mutated using site‐directed mutagenesis using GeneTailor Site‐Directed Mutagenesis System (Invitrogen, Guangzhou, China), yielding the mutant 3′‐UTR of Sema4C (named mt). The wt or mt vector, along with the control vector psiCHECK‐2 vector, were co‐transfected into cells in presence of miR‐25‐3p mimics or siRNA inhibitors in 48‐well plates, and then cells were collected for luciferase assay at 48 h after transfection. Dual‐Luciferase Reporter Assay System (Promega) was used for analysis of luciferase activities, which was normalized to the firefly luciferase activities.

Establishment of stable miR‐25‐3p overexpressing cells

Lentiviral vectors for miR‐25‐3p overexpression were purchased from GeneChem (Shanghai, China). A lentiviral vector that expresses a non‐coding RNA was used as the control. HeLa‐CR cells were seeded at in each well of 24‐well plates (5 × 104 cells in each well) and infected with miR‐25‐3p (Lv‐miR‐25‐3p) or control the lentiviral vector (control LEV) at the multiplicity of infection of 10 (10 infectious units for each target cells). After 72 h of infection, cells were selected with Puromycin and miRNA levels were quantified using qRT‐PCR.

Tumor xenograft model

This study was performed strictly according to the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of The Fourth Clinical Hospital of Harbin Medical University. All surgery was performed when mice were under sodium pentobarbital anesthesia, and efforts were made to avoid animal suffering. Female athymic nude mice of 4–5 weeks were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China) and were maintained in the Animal Resource Facility. Cervical cancer cell line HeLa‐CR, stably transfected with Lv‐miR‐25‐3p or the control LEV, were harvested and suspended in PBS. HeLa‐CR/Lv‐miR‐25‐3p and HeLa‐CR/control LEV (5 × 105) cells were injected subcutaneously in the left and right flank of 4–6‐week‐old female nude mice (n = 5), respectively. Tumors were examined twice a week; length and width were measured with calipers and tumor volumes were calculated using the equation (L × W 2)/2. After 4 weeks, animals were sacrificed, and tumors were collected and embedded in paraffin.

Immunohistochemistry

Tumor sections (5 μm thick) were processed using standard deparaffinization and rehydration protocols. Following rehydration, antigen retrieval was perfomed by submerging the slides in 10 mmol/L sodium citrate buffer (pH 6.0), and heated using a pressure cooker maintained at 95°C for 20 min followed by 20‐min cooling. The sections were then washed with PBS and blocked with 1% BSA/2% goat serum before incubation with either anti‐Ki‐67, anti‐E‐cadherin, anti‐Snail and anti‐Sema4C (Cell Signaling Technology). Subsequently, the sections were incubated with biotinylated secondary antibody followed by HRP‐conjugated streptavidin. The 2,4‐diaminobenzidine (DAB) substrate were then applied to the slides followed by counterstaining with hematoxylin.

Statistical analysis

Data are presented as mean ± SD unless otherwise stated. The statistical significance of the difference between the values of control and treatment groups was analyzed by either Student's t‐test or simple one‐way ANOVA using Prism version 5 (GraphPad Software, Inc., San Diego, CA, USA). Differences with P < 0.05 were considered statistically significant.

Results

CR cells exhibit EMT feature

To determine the underlying mechanism of cervical cancer chemoresistance, we first established the CR cervical cancer cell lines. Higher cell viability was observed in HeLa‐CR and CaSki‐CR cells with cisplatin treatment compared to parental cells (Fig. 1a). The IC50 values derived from cisplatin treated HeLa, HeLa‐CR, CaSki and CaSki‐CR were 9.0 ± 1.2, 79.6 ± 9.8, 10.6 ± 1.4 and 96.5 ± 13.6 nM, respectively. In addition, colony formation assay showed that 10 μM cisplatin significantly inhibited colony formation in HeLa and CaSki cells, while CR cells exhibited resistance to the colony formation inhibitory properties of 10 μM cisplatin (Fig. 1b). These observations indicated that HeLa‐CR and CaSki‐CR cells displayed resistance to cisplatin treatment. The CR cells were exposed to continuous cisplatin selection. It has been demonstrated that drug‐resistant cells possess EMT characteristics.20 The induction of EMT coincided with acquisition of aggressive characteristics, including enhanced cell attachment and cell detachment. Indeed, we found that enhanced attachment and detachment were observed in CR cells (Fig. 1c). These data suggest that CR cells gained a mesenchymal phenotype, which possibly mediated cisplatin resistance in cervical cancer.

Figure 1

Cisplatin‐resistant (CR) cervical cancer cells exhibited epithelial‐mesenchymal transition (EMT) phenotype. (a) Parental (HeLa and CaSki) and CR (HeLa‐CR and CaSki‐CR) cervical cancer cells were treated with increasing concentrations of cisplatin. The cell viability was evaluated by MTT assay. (b) The effect of cisplatin (10 μM) on the growth of CR and parental cells as examined by colony formation assay. Untreated cells were used as the control. (c) Cell attachment and detachment were assessed in parental and CR cells. Data are shown as mean ± SD from triplicate experiments. *P < 0.05, **P < 0.01 versus control.

CR cells display enhanced migratory, invasive ability and induce EMT characteristics

We next examined the migratory and invasive abilities of CR cervical cancer cells, HeLa‐CR and CaSki‐CR, using migration and invasion assays, using parental cells as the control. Wound healing assay revealed that migration of CR cells were higher compared with that of their parental cells (Fig. 2a). Consistently, CR cells have significantly higher number of cells that invaded through the Matrigel‐coated (Fig. 2b). Moreover, to identify the changes of EMT molecular markers in CR cells, the expression of EMT markers in paired parental and CR cell lines were compared using Western blotting analysis. Unsurprisingly, E‐cadherin was significantly decreased in CR cells, whereas mesenchymal markers, such as Snail and Vimentin, were greatly upregulated in CR cells (Fig. 2c,d). These results suggest that CR cervical cancer cells exhibit high ability of cell migration and invasion, along with the acquisition of EMT characteristics.

Figure 2

Cisplatin‐resistant (CR) cells exhibit enhanced motility activity and epithelial‐mesenchymal transition (EMT) marker changes. (a) Wound scratch healing assay was performed in parental (HeLa and CaSki) and CR (HeLa‐CR and CaSki‐CR) cells. (b) Invasion assay was carried out to measure the invasive capacity in parental and CR cells. (c) Western blotting analysis was used to analyze the expression of E‐cadherin, Snail and Vimentin in parental and CR cells. (d) The quantification of Western blotting signal intensities was performed using ImageJ software. Data are shown as mean ± SD from triplicate experiments. *P < 0.05, **P < 0.01 versus control.

Overexpression of miR‐25‐3p reverses EMT in CR cells

It has been demonstrated that miR‐25‐3p level were associated with drug‐resistant in ovarian cancer.19 To define whether miR‐25‐3p is involved in CR cervical cancer cells, we compared the expression of miR‐25‐3p in both CR cells and their parental cells. We observed a significant decrease in miR‐25‐3p level in both HeLa‐CR and CaSki‐CR cells (Fig. 3a). To further determine whether overexpression of miR‐25‐3p could reverse EMT in CR cells, we transfected miR‐25‐3p mimics into HeLa‐CR and CaSki‐CR cells, and found that miR‐25‐3p mimic treatment reduced attachment and detachment capacities in CR cells (Fig. 3b). Moreover, we examined the expression of EMT biomarkers in CR cells transfected with miR‐25‐3p mimics by Western blotting analysis. E‐cadherin expression was significantly increased in CR cells after miR‐25‐3p mimic transfection, whereas the expression of mesenchymal markers, such as Snail and Vimentin, was markedly deceased in CR cells (Fig. 3c,d). These findings identified that reduction of miR‐25‐3p could possibly account for CR‐induced EMT in cervical cancer cells.

Figure 3

miR‐25‐3p mimic reverses epithelial‐mesenchymal transition (EMT) in cisplatin‐resistant (CR) cervical cancer cells. (a) TaqMan miRNA assay was conducted to analyze the expression of miR‐25‐3p in in parental (HeLa and CaSki) and CR (HeLa‐CR and CaSki‐CR) cells. (b, c) HeLa‐CR and CaSki‐CR cells were transfected with miR‐25‐3p mimic and then subjected to cell attachment and detachment assays (b), and Western blotting analysis (c). (d) The quantification of Western blotting signal intensities was performed using ImageJ software. Data are shown as mean ± SD from triplicate experiments. *P < 0.05, **P < 0.01 versus control.

MiR‐25‐3p regulates EMT by targeting Sema4C in CR cells

Since it is widely accepted that miRNA exerts its function via binding to the 3′‐UTR (untranslated region) of target genes through partial sequence homology, we used a further two target prediction programs, TargetScan and miRanda, to screen the potential targets of miR‐25‐3pto define the role of miR‐25‐3p in controlling CR‐induced EMT. Our analysis predicted that Sema4C is a potential miR‐25‐3p target. To confirm this, miR‐25‐3p mimic and Sema4C 3′‐UTR wild type or 3′‐UTR mutated luciferase reporter were simultaneously transfected into CR cells. As a result, transfection with miR‐25‐3p mimics and the reporter plasmid with 3′‐UTR of Sema4C led to a significant decrease in luciferase activity, whereas transfection with miR‐25‐3p mimics and the plasmid without Sema4C 3′‐UTR induced no change in luciferase activity (Fig. 4a). To further validate whether Sema4C served as a direct target of miR‐25‐3p through specific interaction, we measured the mRNA and protein expression of Sema4C after transfecting miR‐25‐3p mimics. Unsurprisingly, Sema4C expression was prominently decreased at both mRNA and protein levels in HeLa‐CR and CaSki‐CR cells under miR‐25‐3p overexpression (Fig. 4b,c). Taken together, miR‐25‐3p specifically targets the 3′‐UTR of Sema4C and subsequently inhibits the expression of Sema4C.

Figure 4

miR‐25‐3p targeted Sema4C expression. (a) Luciferase reporter assays were performed in HeLa‐CR cells to identify the interaction between miR‐25‐3p and Sema4C 3′‐UTR. (b and c) HeLa‐CR and CaSki‐CR cells were transfected with miR‐25‐3p mimics. Real‐time RT‐PCR assay (b) and Western blotting analysis (c) were used to detect the expression of Sema4C in the mRNA and protein levels, respectively. The quantification of Western blotting signal intensities was performed using ImageJ software. (d) CR cells were transfected with miR‐25‐3p mimics or Sema4C siRNA, followed by treatment with 10 μM cisplatin for 72 h. MTT assay was performed to evaluate the cell growth. Data are shown as mean ± SD from triplicate experiments. *P < 0.05, **P < 0.01 versus control.

Upregulation of miR‐25‐3p or depletion of Sema4C enhances CR cells to cisplatin sensitivity

To determine whether upregulation of miR‐25‐3p or depletion of Sema4C enhances cisplatin sensitivity, we performed MTT assay in CR cells after treatment with miR‐25‐3p mimic or Sema4C siRNA. We observed that miR‐25‐3p upregulation or Sema4C downregulation significantly promoted cell growth inhibition induced by 10 μM cisplatin in HeLa‐CR and CaSki‐CR cells (Fig. 4d). These findings suggested that miR‐25‐3p or depletion of Sema4C sensitized CR cells to cisplatin.

Overexpression of miR‐25‐3p inhibits HeLa‐CR xenograft tumor growth

To investigate whether miR‐25‐3p overexpression promotes tumour growth in vivo. HeLa‐CR cells that stably overexpressed miR‐25‐3p were established. A significant increase in miR‐25‐3p level was seen in HeLa‐CR/Lv‐miR‐25‐3p cells compared with that in the HeLa‐CR/control LEV group as revealed by qRT‐PCR (Fig. 5a). Subsequently, HeLa‐CR/Lv‐miR‐25‐3p cells were inoculated subcutaneously in the dorsal flank of nude mice. After 4 weeks of inoculation, the tumors formed by HeLa‐CR/control LEV cells demonstrated significantly larger sizes than HeLa‐CR/Lv‐miR‐25‐3p tumors in volume and weight (Fig. 5b,c). In addition, Ki‐67 staining showed that tumors of HeLa‐CR/control LEV cells exhibited higher proliferative activities compared with HeLa‐CR/Lv‐miR‐25‐3p tumor cells (Fig. 6). Moreover, downregulation of Sema4C and Snail were observed, while upregulation of E‐cadherin in tumors generated from HeLa‐CR/Lv‐miR‐25‐3p cells was seen compared with those generated from HeLa‐CR/control LEV cells, as demonstrated by immunohistochemical analysis (Fig. 6). Taken together, miR‐25‐3p inhibited CR cervical cancer xenograft tumor growth and is associated a reversal of EMT phenotype.

Figure 5

Overexpression of miR‐25‐3p inhibits xenograft tumor growth in vivo. (a) miR‐25‐3p expression in HeLa‐CR/Lv‐miR‐25‐3p and HeLa‐CR/control LEV cells was determined by qRT‐polymerase chain reaction (PCR). Data are shown as mean ± SD from triplicate experiments. (b) Xenograft model in nude mice (n = 5 per group). The HeLa‐CR/Lv‐miR‐25‐3p or HeLa‐CR/control LEV cells were injected subcutaneously into the dorsal flank of the mice. Tumor volumes were measured twice a week. (c) At the end of experiment, tumors weight were measured. Left: images of the tumors from mice in each group. Right: quantification of tumor weight. *P < 0.05, **P < 0.01 versus control.

Figure 6

Immunohistochemistry assay. Left: Representative Ki67, Sema4C, Snail and E‐cadherin staining sections of tumors formed by HeLa/Lv‐miR‐25‐3p or HeLa‐CR/control LEV. Right: quantification of positive cells. For each generated tumor, five fields were randomly selected (n = 5 mice per group). Scale bar, 100 μm. *P < 0.05, **P < 0.01 versus control.

Discussion

Chemoresistance in cancers constitutes the main cause of treatment failure.21 Although recent data indicated that miRNA dysregulation is closely associated with chemoresistance by post‐transcriptionally regulating genes linked to chemosensitivity22 or chemoresistance,23 the specific miRNAs that specifically regulate chemoresistance are largely unknown. More studies are warranted to identify miRNAs involved in chemoresistance and the underlying mechanisms by which they induce chemoresistance. Cisplatin has been used to treat cervical cancer since the early 1980s24 and remains the most effective anticancer agent for advanced/recurrent cervical cancer. However, chemoresistance that emerges in cervical cancer may seriously hinder the clinical value of cisplatin. Therefore, there is an urgent need to delineate the underlying mechanism of cisplatin resistance in cervical cancer. Therefore, in this study, we investigated the molecules that are associated with cisplatin resistance in cervical cancer. We demonstrated that CR cervical cancer cells are characterized by EMT, through which cells gained increased migration and invasion. In addition, downregulation of miR‐25‐3p was observed in CR cells, and overexpression of miR‐25‐3p led to the reversal of EMT phenotype and sensitized CR cells to cisplatin through targeting Sema4C. Moreover, ectopic expression of miR‐25‐3p in HeLa‐CR cells inhibited xenograft tumor growth in mice. This was concomitant with downregulation of Sema4C and Snail, as well as upregulation of E‐cadherin compared with the control group. Our results indicate that miR‐25‐3p is associated with chemoresistance and induction of EMT characteristics in cervical cancer cells. Therefore, upregulation miR‐25‐3p could be a novel strategy for overcoming drug resistance of cervical cancer.

A large body of literatures strongly suggested the link between EMT and chemoresistance in cancer. Previous study suggested that pancreatic cancer cells gained gemcitabine‐resistance by undergoing EMT through Notch signaling pathway.25 Consistently, gemcitabine‐resistant hepatocellular carcinoma cells exhibited EMT characteristics, such as E‐cadherin downregulation and Vimentin, Snail, and Slug upregulation.26 Paclitaxel‐resistant ovarian cancer cells also displayed EMT phenotype and high expression of PI3K.27 On the other hand, knockdown of astrocyte elevated gene‐1 blocked EMT and reduces cisplatin resistance in cervical cancer cells.28 Our study mirrored these reports, and demonstrated that CR cervical cancer cells had an EMT phenotype, characterized by the downregulation of E‐cadherin and upregulation of Snail and Vimentin. Phenotypically, CR cervical cancer possessed enhanced migration and invasion. Our results suggest that, in cervical cancer, cisplatin resistance is associated with EMT.

Recent studies have highlighted the important role of miRNAs in EMT resulting from chemotherapy. For instance, the miR‐134/487b/665 cluster was shown to govern transforming growth factor‐β (TGF‐β)‐mediated EMT and drug resistance to gefitinib by targeting membrane associated guanylate kinase inverted 2 (MAGI2) in cervical cancer cells.29 In addition, it also has been reported that miR‐489 modulated chemoresistance via EMT‐related pathway in breast cancer.30 Moreover, Shien et al.31 unraveled that acquisition of resistance to EGFR inhibitors stemmed from EMT characteristic of miR‐200c downregulation in cancer cells. Interestingly, miR‐216a/217‐induced EMT‐stimulated drug resistance by targeting phosphatase and tensin homolog (PTEN) and mothers against decapentaplegic homolog 7 (SMAD7) in hepatocellular carcinoma.32 In support of the role of miRNAs in regulating chemotherapy‐induced EMT, we observed that miR‐25‐3p was downregulated in CR cells. Our finding is consistent with previous report that miR‐25‐3p was downregulated in drug‐resistant ovarian cancer.19 Additionally, lower miR‐25 levels are reported in colorectal cancer tissues compared with those in matched non‐tumor tissues, and restoring miR‐25 expression suppresses the cell proliferation and migration of colorectal cancer cells.33 In contrast, higher levels of miR‐25 have been observed in gastric cancer tissues and cell lines, papillary thyroid carcinomas, lung adenocarcinoma, malignant cholangiocarcinoma, breast cancer, esophageal squamous cell carcinoma and HBV‐positive hepatocarcinoma.17, 34, 35, 36, 37, 38, 39 It is worth noting that overexpression of miR‐25‐3p or depletion of Sema4C sensitized CR cervical cancer cells to cisplatin. Several studies have demonstrated that Sema4C, a member of the semaphorin family, is crucial in regulating EMT. For example, Sema4C induced EMT through suppressing E‐cadherin and induction of Vimentin in HK2 cells.40 In line with this, Sema4C knockdown reversed TGF‐β1 (transforming growth factor, β1)‐mediated EMT through suppressing the phosphorylation of p38 MAPK (mitogen‐activated protein kinase), whereas Sema4C overexpression promoted EMT via activation of p38 MAPK phosphorylation in human proximal tubular epithelial cells.41 Furthermore, upregulation of Sema4C has been observed in esophageal cancer, gastric cancer and rectal cancer and was shown to be correlated with lymphatic metastasis.42 In support of the role of Sema4C in controlling EMT, our results showed that miR‐25‐3p modulated CR‐induced EMT in part due to downregulation of its target Sema4C. Furthermore, stable overexpression miR‐25‐3p in HeLa‐CR cells inhibited tumor xenograft growth in immunodeficient mice, along with inhibition of cell proliferation, suppression of Sema4C and Snail expression, and upregulation of E‐cadherin in tumor xenograft tissues. However, recent studies showed that miR‐25 is remarkably increased in lung cancer tissue samples and cell lines. Downregulation of miR‐25 significantly suppressed lung cancer cell proliferation, migration, invasion and the lung cancer cell xenograft tumor growth in vivo.43, 44 These results suggest that miR‐25 may play different role in different type of cancer cells. Taken together, our results suggest that miR‐25‐3p is responsible for the chemoresistance and EMT characteristics of cervical cancer cells in vitro and in vivo. Of course, in‐depth investigation is necessary to explore the function of miR‐25‐3p in different tumor entities.

In summary, our data suggest that CR cells went through EMT partly by way of miR‐25‐3p downregulation. More importantly, overexpression of miR‐25‐3p reversed CR‐induced EMT and sensitized CR cells to cisplatin by targeting Sema4C. This study indicated that activation of miR‐25‐3p could be a novel strategy to combat chemotherapy resistance in cervical cancer.

Disclosure Statement

The authors declare that there is no conflict of interests.