Role of miR-337-3p and its target Rap1A in modulating proliferation,invasion,migration and apoptosis of cervical cancer cells

Abstract

OBJECTIVE:

To investigate the role of miR-337-3p targeting Rap1A in modulating proliferation, invasion, migration and apoptosis of cervical cancer cells.

METHODS:

The expression levels of miR-337-3p and Rap1A in cervical cancer tissues and normal tissues were evaluated through quantitative Real-time PCR (qRT-PCR) and Western blotting; and correlations of miR-337-3p with clinicopathological characteristics and prognosis of patients were also analyzed. Besides, human cervical cancer cell line HeLa cells were randomly divided into five groups (Mock, NC, miR-337-3p mimic, Rap1A, and miR-337-3p mimic $+$ Rap1A groups). CCK-8 assay was utilized to measure cell proliferation, flow cytometry to evaluate cell apoptosis, and wound-healing and Transwell assays to detect cell migration and invasion.

RESULTS:

Cervical cancer tissues presented a significant decrease in miR-337-3p and a remarkable increase in Rap1A protein. Besides, the expression levels of miR-337-3p and Rap1A were closely related to the major clinicopathological characteristics of cervical cancer; and patients with high-miR-337-3p-expression had the higher 5-year survival rate (all $p<$ 0.05). When compared to Mock group, cells in miR-337-3p mimic group were suppressed in proliferation, migration, and invasion, but significantly promoted in apoptosis; meanwhile, cells in the Rap1A group showed changes in a completely opposite trend (all $p<$ 0.05). Moreover, Rap1A can reverse the effect of miR-337-3p mimic on cell proliferation, invasion, migration and apoptosis (all $p<$ 0.05).

CONCLUSION:

MiR-337-3p was discovered to be decreased in cervical cancer, and miR-337-3p up-regulation may inhibit the proliferation, migration and invasion and promote the apoptosis of cervical cancer cells via down-regulating Rap1A.

1. Introduction

Cervical cancer, as a malignant tumor found in the cervical epithelium, is one of the most common gynecologic malignancies with very high incidence [1]. According to the statistics, there are over 500,000 newly diagnosed cervical cancer patients and nearly 250,000 deaths caused by cervical cancer each year, posing a serious threat to the health of women worldwide [2, 3]. Unfortunately, the current major therapies for cervical cancer including surgery, radiotherapy, chemotherapy or other comprehensive treatments, are very limited in the curative effect for patients with advanced cervical cancer [4], which had an important bearing on the invasion and metastasis of cervical cancer [5]. As documented previously, cervical cancer is suggested to be an extremely complex process, including the hyperplasia of normal cervical epithelium, the metaplasia to cervical intraepithelial neoplasia (CIN) I-III, and eventually the carcinogenesis, which is a multi-step process regulated by multiple genes [6, 7]. Besides, high-risk human papillomavirus (HR-HPV) infection has been recognized as one of the most important risk factors of cervical cancer [8]. Currently, the HPV-vaccines available protect against the two high-risk HPV types (including 16 and 18) that induce most cervical cancer [9], and cervical cancer caused by HPV might be successfully prevented by HPV vaccination and screening to some extent [10]. However, HPV vaccines are designed for prophylactic use only, which must be given before the infection occurs [11]. They do not clear existing HPV infection or treat HPV-related disease such as cervical cancer [12]. Therefore, we try best to explore in-depth the pathogenesis of cervical cancer in this study, hoping to lay a theoretical foundation for the early diagnosis and effective treatment of cervical cancer.

MicroRNAs (miRNAs), a class of non-coding endogenous single-strand RNA molecules composed of approximately 20–24 nucleotides, are widely distribu- ted in eukaryotes, which could control tumors’ formation and progression at different time points via regulating various cellular activities, such as proliferation, metastasis, and angiogenesis [13, 14]. In recent years, accumulating evidence has shown that many miRNAs were abnormally expressed in cervical cancer, such as miR-101 [15], miR-497 [16], and miR-372 [17], demonstrating significant effects of miRNAs on the early diagnosis and therapy selection for cervical cancer patients. Hsa-miR-337 (miR-337), a human miRNA locus located at chromosome 14q32.2 [18], has been reported to suppress the proliferation and invasion of pancreatic cancer through down-regulation of HOXB7 in the study by Zhang et al. [19], which could also inhibit the development of neuroblastoma by specifically reducing the expression of MMP-14 [20]. Of note, miR-337-3p could target both STAT3 and Rap1A, to modulate the progression and drug resistance of non-small cell lung cancer [18]. As for Rap1A, it is a member of the Ras-associated proteins (Raps) with the highest resemblance to Ras [21] and its function was realized by modulating endothelial cell junctions through E-cadherin, which was thought to be associated with the pathogenesis and development of tumors [22]. More importantly, the over-expression of RAP1 has an important bearing on the cervical intraepithelial neoplasia [23]. By referring to the target gene prediction website, we predicted Rap1A to be the target gene of miR-337-3p. However, no relevant report has elucidated clearly whether miR-337-3p can affect the development of cervical cancer through the modulation of Rap1A. Therefore, in this study, we investigated the effect of miR-337-3p and Rap1A on the proliferation, invasion, migration and apoptosis of cervical cancer cells, for the purpose of providing some new therapeutic targets for the diagnosis and treatment of cervical cancer.

2. Materials and methods

2.1 Ethics statement

All patients in the present study were informed of the experimental content and signed the written informed consent form. And the study was approved by the Ethics Committee for Clinical Experiments in The First People’s Hospital of Jingzhou City.

2.2 Subjects of the study

From January 2015 to January 2017, 86 cases of cervical cancer patients were enrolled as Case group of the study, who were 31–62 years old with the mean age of 48.7 $\pm$ 9.0 years. Before the operation, none of the patients received either radiotherapy or chemotherapy, and they didn’t have any malignant tumors of other organs. Based on the classification criteria provided by the International Federation of Gynecology and Obstetrics (FIGO), there were 49 patients at the Ib-IIa phase and 37 cases at the IIb-IIIa phase. Among them, 48 cases were with well-moderate differentiation and 38 cases with poor differentiation; and 32 cases with lymphatic metastasis and 54 cases without metastasis. During the same time period, normal cervical tissues were obtained from 25 patients who were about matched ages of the Case group and received total hysterectomy for hysteromyoma, thereby constituting the Control group. Cervical intraepithelial neoplasia (CIN) tissues, which were classified histopathologically into 35 cases of CIN I stage, 28 cases of CIN II phase and 32 cases of CIN III stage, were also collected. All specimens were obtained from the fresh isolated cervix or biopsy specimens of living cervix, which were stored at $-$ 80 ${}^{\circ }$ C for later use.

2.3 Follow-up

The initial time point of follow-up was the date of diagnosis and the survival period was calculated on a monthly basis. The follow-up endpoint was patient death. Besides, follow-up examination was usually carried out every 3 months for the first year after treatment, every 3 to 6 months for the 2nd year after treatment, every 6 months for the 3rd year after treatment, and once a year since the 5th year after treatment.

2.4 Culture and transfection of cells

HeLa cells (the human cervical cancer cell line) were purchased from the American Type Culture Collection (ATCC) and then cultured in the RPMI-1640 medium (Hyclon) supplemented with 10% fetal bovine serum (FBS, Gibco), 100 units/ml penicillin (Hyclon) and 100 mg/ml streptomycin (Hyclon), in an incubator with 5% CO ${}_2$ at 37 ${}^{\circ }$ C (Thermo). When cells covered 80% of the visual field under the microscope, they were digested by 0.25% trypsin for passaging and subsequent experiments. HeLa cells were spread on the plate 24 h before transfection, and a final concentration of 50 nM miR-337-3p mimics/NC or 20 $\mu$ M of Rap1A was transfected into the cells in accordance with the instructions on the Lipofectamine 2000 Kit (Invitrogen). In the experiment, HeLa cells were grouped as follows: Mock group (HeLa cells without transfection), NC group, miR-337-3p mimic group, Rap1A group and miR-337-3p mimic $+$ Rap1A group. Above all, miR-337-3p mimics, Negative control (NC) sequences, and Rap1A-over-expression plasmids were bought from Shanghai GenePharma Co., Ltd.

2.5 Dual-luciferase reporter gene assay

In order to explore the impact of miR-337-3p on the expression of Rap1A, human Rap1A 3’-UTR fragment located at 3’-UTR and with the binding site of miR-337-3p was amplified to construct the wild-type plasmid Rap1A WT and the mutant-type plasmid Rap1A MUT. Cells were transfected with Lipofectamine 2000 Kit (Invitrogen) and the transcriptional activity was determined by the dual-luciferase reporter gene assay kit (Promega) according to the manufacturer’s instructions. The co-transfection systems were as follows: miR-337-3p mimic $+$ Rap1A-WT, mimic-NC $+$ Rap1A-WT, miR-337-3p mimics $+$ Rap1A-MUT, and mimic-NC $+$ Rap1A- MUT.

2.6 Quantitative Real-time PCR (qRT-PCR)

Total RNA was extracted from tissues and cells with TRIzol Kit (Invitrogen), determined for concentration with NanoDrop2000 reagent (Thermo), and preserved at $-$ 80 ${}^{\circ }$ C for later use. The Transcriptor First Strand cDNA Synthesis Kit (Roche, Indianapolis, IN) was applied for the reverse transcription of RNA to cDNA, and qRT-PCR was performed by using the SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA). U6 was used as the endogenous control and 2 ${}^{-Ct}$ method to calculate the relative expression level of target genes, with the formulae: Ct $=$ CT(target gene)-CT(internal reference gene) and Ct $=$ Ct(treatment group)-Ct(control group). All experiments were performed in triplicate.

2.7 Western blotting

Proteins extracted from tissues and cells were quantified by the BCA kit (Thermo Fisher, Rockford, IL). Then, proteins were heated for 10 min at 95 ${}^{\circ }$ C after the addition of loading buffer by 30 $\mu$ g/well. And 10% polyacrylamide gel (Wuhan Boster Biological Technology Co., Ltd.) was used to resolve proteins by electrophoresis, which was transferred onto the poly vinylidene fluoride (PVDF) membrane before 1 h of blocking with 5% bovine serum albumin (BSA) at room temperature. Next, the membranes were probed with primary antibodies including Rap1A (1:500 diluted, ab115776, Abcam) and $\beta$ -actin (1:1000 diluted, ab8226, Abcam) overnight at 4 ${}^{\circ }$ C. After that, the membrane was washed with TBST for 3 times/5 min and incubated with secondary antibody (1:2000 diluted) for 1 h at room temperature. Later, the membrane was rinsed again for 3 times/5 min before development by using the chemiluminescence reagent, with $\beta$ -actin as the loading control. All experiments were performed in triplicate.

2.8 Serum-free cell survival assay

2 $\times$ 10 ${}^5$ cells were seeded and the number of living HeLa cells was determined by Countess automated cell counter (Invitrogen) after treating them in serum-free or 10% fetal bovine serum DMEM for 24, 48 and 72 hours.

2.9 Cell proliferation by CCK-8 assay

HeLa cells were plated into the 96-well plate, with 100 $\mu$ l cell suspension in each well (5000 cells/well). At the 24 h, 48 h and 72 h after transfection, the Cell Counting Kit-8 (CCK-8) (Beyotime Institute of Biotechnology) was applied for the detection of cell proliferation. According to the instructions, 10 $\mu$ l of CCK-8 reagent was added into each well of the plate for 3 h of incubation at 37 ${}^{\circ }$ C. After that, the absorbance (OD value) of each well was measured on a multi-functional microplate reader (Thermo) at the wavelength of 490 nm. All experiments were performed in triplicate.

2.10 Cell apoptosis by flow cytometry

HeLa cells were digested by trypsin, centrifuged, and collected, which were washed with ice-cold PBS buffer and re-suspended into single cell suspension (with 1 $\times$ 10 ${}^6$ cells/ml) with the use of buffer solution (calcium containing PBS buffer). Next, 100 $\mu$ l cell suspension was placed into the tube at room temperature, followed by incubation of propidium iodide (with final concentration of 10 mg/ml) and RNase A (10 mg/ml) for 30 min at 4 ${}^{\circ }$ C. Then, after the addition of 400 $\mu$ l binding buffer, cell suspension was loaded onto the flow cytometer (BD Bioscience) for detection and analysis (with 10 ${}^4$ cells each time). The software Cell Quest was applied for data analysis and the calculation of cell apoptosis rate. All experiments were performed in triplicate.

2.11 Wound-healing assay

Transfected HeLa cells were seeded into 6-well plates for incubation at 37 ${}^{\circ }$ C. When cells covered the bottom of the plate, the sterilized tip of transfer liquid gun was used to gently draw lines on the plate and the width of all scratches should be about the same, which was photographed before covering the 6-well plate (0 h). Then, cells were incubated for 48 h at 37 ${}^{\circ }$ C, followed by removing cell culture fluid and washing the plate with PBS buffer for 3 times to wash away cell debris produced by scratches. Next, serum-free medium was added into the plate for blocking cell proliferation, which was photographed again (48 h). The OLYMPUS inverted microscope was used to take photos of 6 randomly selected visual fields and the number of cells migrating to the scratched area was counted by using click-counting software. All experiments were performed in triplicate.

2.12 Transwell invasion assay

After 48 h transfection, matrigel (3.9 mg/ml, 60–80 $\mu$ l) was added into Transwell chambers (German Chambers), which was incubated for 30 min at 37 ${}^{\circ }$ C for gel solidification. Before the experiment, pre-heated medium should be added into the lower and upper chambers for hydration. After 2 h of hydration at 37 ${}^{\circ }$ C, the plate was taken out and 0.5 ml complete medium was added into each well, followed by the addition of 0.5 ml cell suspension (5 $\times$ l0 ${}^4$ cells/ml) to the upper chamber for 24 h of incubation at 37 ${}^{\circ }$ C. Next, the chamber was taken out to wipe away cells on the upper chamber, followed by washing with PBS buffer and fixation for 30 min with pre-cooled methanol. Six randomly selected visual fields were observed and photographed by using the Olympus inverted microscope, and the number of cells migrating to the lower chamber was counted by using the click-counting software. All experiments were performed in triplicate.

2.13 Statistical method

Statistical analysis was performed using SPSS 21.0 software. Measurement data were presented by mean $\pm$ standard deviation ( $\overline{x}$ $\pm$ s). Two groups of measurement data conforming to the normal distribution were compared by using student’s t-test, while multiple groups of data were analyzed by using One-Way ANOVA. Meanwhile, enumeration data were presented by percentage and ratio, and tested by Chi-square test ( ${\chi }^2$ ). Overall survival curves were plotted according to the Kaplan-Meier method with the log-rank test for comparison. $p<$ 0.05 indicated the statistical significance.

3. Results

3.1 Expression of miR-337-3p and Rap1A in cervical tissues

The results of qRT-PCR demonstrated that miR-337-3p was markedly down-regulated in cervical cancer tissues when compared with normal tissues ( $p<$ 0.05, Fig. 1A). The results were shown in Fig. 1A that the expression of miR-337-3p in the samples of CIN I and CIN II stage did not change significantly compared with that in normal cervical tissues (both $p>$ 0.05), whereas decreased significantly in the samples of CIN III stage (all $p<$ 0.05). What’s more, miR-337-3p was decreased apparently in cervical cancer tissues when compared with CIN III stage tissues ( $p<$ 0.05). Besides, according to the results of Western blotting, the expression of Rap1A protein in cervical cancer tissues was remarkably higher than that in normal tissues ( $p<$ 0.05, Fig. 1B and C).

Table 1

The association of miR-337-3p expression with cervical cancer patients’ clinicopathological features

Characteristics	Cases	MiR-337-3p expression	$p$
Age (year)			0.938
$⩾$ 50	46	0.520 $\pm$ 0.120
$<$ 50	40	0.522 $\pm$ 0.118
Tumor size (cm)			0.906
$<$ 4	39	0.517 $\pm$ 0.115
$⩾$ 4	47	0.520 $\pm$ 0.119
FIGO stage			$<$ 0.001
Ib-IIa	49	0.593 $\pm$ 0.092
IIb-IIIa	37	0.424 $\pm$ 0.139
Differentiation			$<$ 0.001
Well-moderately differentiated	48	0.595 $\pm$ 0.092
Poor differentiation	38	0.426 $\pm$ 0.077
Lymph node metastasis			$<$ 0.001
Yes	32	0.415 $\pm$ 0.151
No	54	0.583 $\pm$ 0.094

Figure 1.

Expression levels of miR-337-3p and Rap1A in cervical cancer tissues and normal tissues. Note: A: The expression of miR-337-3p in cervical cancer tissue, cervical intraepithelial neoplasia (CIN) tissues and normal tissues determined by qRT-PCR; B, C: The expression of Rap1A protein in cervical cancer tissues and normal tissues evaluated by Western blotting; D: The effect of miR-337-3p expression on the 5-year survival rate of cervical cancer patients.

Figure 2.

The targeting relationship between miR-337-3p and Rap1A. Note: A, The binding site of miR-337-3p to Rap1A predicted by using a public database and the results of dual-luciferase reporter gene assay; B, The expression of miR-337-3p in each group detected by qRT-PCR; C, D: The expression of Rap1A protein determined by Western blotting; *, $p<$ 0.05 compared with the Mock group; ${}^{\mathrm{\#}}$ , $p<$ 0.05 compared with the miR-337-3p mimic group.

3.2 Association of miR-337-3p with cervical cancer patients’ clinicopathological features and prognosis

As shown in Table 1, the expression level of miR-337-3p in cervical cancer tissues had no correlation to the age and tumor size of patients (both $p>$ 0.05), but were closely correlated with FIGO stage, differentiation degree, and lymph node metastasis. In other words, patients with poorer differentiation, higher FIGO stage, and lymph node metastasis had lower expression of miR-337-3p ( $p<$ 0.05). Patients were divided into two groups based on miR-337-3p expression levels (low vs. high) with the median expression level as a cutoff point to analyze the effect of miR-337-3p on the prognosis of patients. Accordingly, the cervical cancer patients with high-miR-337-3p-expression were significantly elevated in survival rate, and the 5-year survival rate was statistically higher than those with low-miR-337-3p-expression ( $p=$ 0.014, Fig. 1D).

Figure 3.

The impact of miR-337-3p on the proliferation and apoptosis of HeLa cells in each transfection group. Note: A, Survival of HeLa cells in serum-free or 10% fetal bovine serum medium; B, The impact of miR-337-3p expression on the HeLa cell proliferation in each group detected by CCK-8 assay; C, The impact of miR-337-3p expression on the HeLa cell apoptosis detected by flow cytometry; *, $p<$ 0.05 compared with the Control group; ${}^{\mathrm{\#}}$ , $p<$ 0.05 compared with the miR-337-3p mimic group.

3.3 MiR-337-3p could target Rap1A

According to the prediction on the microRNA.org, the putative binding sequence of miR-337-3p was located on the 3’-UTR of Rap1A (Fig. 2A). The luciferase reporter gene assay showed that the luciferase activity was significantly decreased in cells co-transfected with the Rap1A WT and miR-337-3p mimic (all $p<$ 0.05); however, in the Rap1A MUT co-transfection system, no significant difference was found regarding luciferase activity, which suggested that Rap1A was the target gene of miR-337-3p. By detecting the expression of miR-337-3p and Rap1A protein in each transfected group of cells, no differences were found in the expression of miR-337-3p and Rap1A between Mock and NC groups (all $P>$ 0.05). Besides, miR-337-3p mimic group was remarkably elevated in miR-337-3p, but clearly reduced in Rap1A protein (all $p<$ 0.05); and meanwhile, Rap1A group presented no significant difference in the expression of miR-337-3p (all $p>$ 0.05), but had remarkable increase in the expression of Rap1A, as compared with the Mock group (all $p<$ 0.05). Moreover, miR-337-3p mimic $+$ Rap1A group had no significant change in the expression in miR-337-3p, but were dramatically elevated in Rap1A when compared to the miR-337-3p mimic group (Fig. 2B–D).

Figure 4.

The impact of miR-337-3p on the migration and invasion of HeLa cells. Note: A, The migration ability of HeLa cells in each group evaluated by the wound-healing assay; B, The invasion ability of HeLa cells in each group measured by Transwell Invasion assay; *, $p<$ 0.05 compared with the Control group; ${}^{\mathrm{\#}}$ , $p<$ 0.05 compared with the miR-337-3p mimic group.

3.4 Impact of miR-337-3p on the proliferation and apoptosis of cells

To investigate the effects of serum-free environment on the proliferation of HeLa cells, the cells were assessed cell proliferation over a three-day period. Although 10% fetal bovine serum-treated cells showed an exponential increase in cell count, cells treated with serum-free environment did not show much change in the number of viable cells, even after 3 days of treatment (Fig. 3A). Our data indicated that serum-free environment was able to rapidly inhibit proliferation of HeLa cells. Meanwhile, CCK-8 assay and flow cytometry were employed to evaluate the proliferation and apoptosis of cells, respectively (Fig. 3B and C). Compared with cells in the Mock group, those in the miR-337-3p mimic group were decreased in proliferation and elevated in apoptosis rate (both $p<$ 0.05); however, cells in the Rap1A group presented completely opposite tendency in cell proliferation and apoptosis (both $p<$ 0.05); meanwhile, no significant difference was observed in cells of the NC group with respect to cell proliferation and apoptosis (both $p>$ 0.05). Moreover, with miR-337-3p mimic group as the baseline for comparison, the miR-337-3p mimic $+$ Rap1A group was significantly promoted in cell proliferation and dramatically reduced in cell apoptosis (both $p<$ 0.05).

3.5 Impact of miR-337-3p on the migration and invasion of HeLa cells

The migration and invasion abilities of HeLa cells were measured by wound-healing and Transwell assays separately. Obviously, miR-337-3p mimic group was statistically lower than the Mock group concerning cell migration and invasion abilities (both $p<$ 0.05), while Rap1A group was significantly higher than the Mock group as regards cell migration and invasion abilities (both $p<$ 0.05). Moreover, in comparison with miR-337-3p mimic group, miR-337-3p mimic $+$ Rap1A group exhibited increased migration and invasion abilities (both $p<$ 0.05, Fig. 4).

4. Discussion

In this study, we detected the expression levels of miR-337-3p and Rap1A in cervical cancer tissues and normal tissues, and found the remarkable down-expression of miR-337-3p and the appreciable up-expression of Rap1A in cervical cancer tissues, which suggested that miR-337-3p may serve as a tumor suppressor gene while Rap1A as an oncogene in cervical cancer. What’s more, miR-337-3p was found to be down-regulated in a variety of tumors, including gastric cancer [24], non-small cell lung cancer [18], and pancreatic cancer [25]. On the other hand, Rap1, classified into two subtypes, namely Rap1A and Rap1B, was widely distributed in cells, which can be converted to inactive GDP-bound form or active GTP-bound form in a dynamic process [26]. Rap1-GTP as the active form was found abnormally expressed in many malignant tumor tissues [22, 27]. Evidence has pointed out that Rap1A expression was highly elevated in gynecologic malignancies, such as breast cancer and ovarian cancer [28, 29]. Worth mentioning, the persistent human papillomavirus (HPV) infection was the major risk factor of cervical cancer [21], and the HR-HPV E6 on its early protein encoding region can specifically degrade E6TP1 (which was one type of GTPase-activator proteins, RapGAPs) [30], to enhance the GTP hydrolysis activity of Rap1, thereby leading to the over-expression of Rap1 [30]. Recently, another study reported that Rap1-GTPase expression was significantly increased in cervical intraepithelial neoplasia and cervical cancer tissues with the severity of cervical lesions [23], providing the possibility that the increase of Rap1A expression in cervical cancer may result from its activation by HPV in the process of cervical carcinogenesis. Moreover, our study also found that miR-337-3p was lower in cervical cancer patients with poor differentiation, higher FIGO staging, and lymph node metastasis, as well as a poorer prognosis. Similarly, Xiang et al. also observed neuroblastoma patients had a better prognosis with a relatively higher expression of miR-337-3p, which could inhibit the disease progression in vitro and in vivo through down-regulation of MMP-14 [20]. The afore-mentioned findings all added weight to the hypothesis that miR-337-3p was closely associated with the pathogenesis and development of cervical cancer, and it was expected to become a biomarker in the prediction of patients’ prognosis.

Furthermore, it has been well-recognized that Rap1 can act as the target protein of some miRNAs to participate in the regulation of tumor progression. For example, a previous study reported that Rap1 was regulated by miR-518b in esophageal squamous cell carcinoma (ESCC) and the exogenous over-expressed Rap1 can reverse the regulatory role of miRNA [31]. In colorectal cancer, silencing miR-139 could suppress the tumor growth and carcinogenicity by up-regulating the expression of its target gene RAP1B [32]. Besides, Lin et al. [33] observed lowly-expressed miR-708 and highly-expressed Rap1B in ovarian tumors, and they also proved that miR-708 could down-regulate Rap1B in ovarian cancer cell lines to modulate the invasion and migration of cancer cells. Another important finding of our study was that Rap1A was the target gene of miR-337-3p, which was confirmed by the dual-luciferase reporter gene assay and further proved that miR-337-3p could specifically act on Rap1A in cervical cancer to affect the biological characteristics of cancer cells. On the other hand, in vitro cell transfection experiments in this study further proved that over-expressed miR-337-3p could restrain the proliferation, migration and invasion of cervical cancer cells, and enhance cell apoptosis rate at the same time. Meanwhile, the over-expression of Rap1A resulted in the completely opposite results. Consistent with our results, miR-337-3p could specifically down-regulate the expression of Rap1A to affect the development of non-small cell lung cancer, as indicated by Du et al. [18]. Another important finding of this study was that over-expressed Rap1A could reverse the inhibitory role of miR-337-3p in cell growth, which further added weight to the hypothesis that miR-337-3p played its roles in cervical cancer by targeting Rap1A. More similar examples have been presented in recent studies. For instance, it has been reported that Rap1 can promote the invasion and distant metastasis of human pancreatic cancer and breast cancer by mediating integrin and regulating actin reorganization [34, 35]. Besides, Rap1A could activate the transcription of MMP-7 in squamous cell carcinoma of the head and neck via the mediation of integrin and MMP-7 acted as an important biomarker for the invasion and migration of malignant cells [36]. Additionally, Rap1 could not only promote cell migration and inhibit cell differentiation by activating the MAPK/ERK signaling pathway, but also promote integrin-dependent cell adhesion and change cell invasion and metastasis [37, 38]. Moreover, Rap1-GTP could induce Rac1 to convert to active GTP-binding form, which can interact with the downstream effector PAK1 and activate the Rac1/PAK1 signaling pathway that was associated with the migration, invasion and other cellular processes [39]. Based on those findings, we hypothesized that miR-337-3p may target Rap1A to modulate the expression of its downstream genes and pathways, and thereby regulate the proliferation, migration and invasion of cervical cancer cells.

Taken together, our study found decreased miR-337-3p expression in cervical cancer tissues. Besides, over-expressed miR-337-3p may inhibit the proliferation, migration and invasion of cervical cancer cells and promote cell apoptosis by reducing the expression of Rap1A. Therefore, miR-337-3p may become a new diagnostic and prognostic biomarker for cervical cancer, and it also provides a novel perspective for the finding of therapeutic targets.