Paired cervical cancer and paracancerous tissues were obtained from 54 patients with cervical cancer who were hospitalized in Shangluo Central Hospital (Shangluo, China) from 2015 to 2017. Before analysis, all specimens were frozen in liquid nitrogen immediately after surgery and stored at −80°C. For this cohort, 30 patients were diagnosed at early stage (0/I/II), while the others were diagnosed at advance stage (III/V), according to the International Federation of Gynecology and Obstetrics (FIGO). Stage grouping and the detailed clinical information are shown in Table I. None of the patients had undergone chemotherapy or radiotherapy before surgery. For all specimens informed consent was obtained from the patients and the study was approved by the Ethics Committee of Shangluo Central Hospital.

Human cervical cancer cell lines HeLa (cat. no. CCL-2), JAR (cat. no. HTB-144), and normal cervical immortalized squamous cells Ect1/E6E7 (cat. no. CRL-2614) were purchased from American Type Culture Collection (ATCC; Rockville, MD, USA). The cells were cultured in a cell incubator at 37°C using RPMI-1640 medium (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc., Carlsbad, CA, USA). Both media were supplemented with penicillin-streptomycin (final concentration of penicillin was 100 U/ml and of streptomycin was 0.1 mg/ml; Beijing Solarbio Science & Technology Co., Ltd., Beijing, China).

Total RNA and total miRNA were extracted utilizing TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and miRcute miRNA isolation kit (Tiangen Biotech Co., Ltd., Beijing, China), respectively. After measuring concentration, PrimeScript™ II 1st Strand cDNA Synthesis kit (Takara Biotechnology Co., Ltd., Dalian, China) was used to synthesize cDNA. In addition, SYBR Premix kit and SYBR PrimeScript miRNA RT-PCR kit (both from Takara Bio, Inc., Otsu, Japan) were employed to perform qPCR. The primer sequences used were: miR-26b forward, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACGAGCCA-3′ and reverse, 5′-CGCCCTGTTCTCCATTACTT-3′; JAG1 forward, 5′-ATCGTGCTGCCTTTCAGTTT-3′ and reverse, 5′-GATCATGCCCGAGTGAGAA-3′; GAPDH forward, 5′-CCACTCCTCCACCTTTGAC-3′ and reverse, 5′-ACCCTGTTGCTGTAGCCA-3′; and U6 forward, 5′-CTTGGCAGCACATATACT-3′ and reverse, 5′-AAAATATGGAACGCTTCACG-3′. The thermocycling conditions were as follows: 2 min at 95°C, followed by 40 cycles of 30 sec at 95°C and 45 sec at 60°C. The expression levels were quantified using the 2^−∆∆Cq method (20). GAPDH and U6 were utilized to normalize mRNA and miRNA, respectively.

Specific cells were lysed by RIPA lysis buffer with 1% protease inhibitor phenylmethanesulfonyl fluoride (PMSF; Beyotime Institute of Biotechnology, Shanghai, China) to obtain total proteins. After quantified with BCA Protein Assay kit (Beijing Solarbio Science & Technology Co., Ltd.), 50 µg of proteins were added into each polyacrylamide gel electrophoresis well and electrophoresis was carried out to separate all proteins. Then, the blots were transferred onto a polyvinylidene fluoride (PVDF) membrane (Bio-Rad Laboratories, Inc., Hercules, CA, USA) and blocked with 5% skim milk powder at room temperature for 1 h. The membrane was incubated with rabbit JAG1 polyclonal antibody (cat. no. PAB807Mu01; 1:500; Wuhan USCN Business Co., Ltd., Wuhan, China) and mouse GAPDH momoclonal antibody (cat. no. sc-32233; 1:2,000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) at 4°C overnight. Then, the membrane was incubated with rabbit antibody labeled with HRP (cat. no. 5571; 1:5,000; Cell Signaling Technology, Inc., Danvers, MA, USA) as secondary antibody for 2 h at room temperature. ECL Western Blotting Detection System (BestBio, Beijing, China) was employed to analyze the protein bands.

The capacity of cell migration and invasion were measured with Transwell assay. The Transwell chambers with or without Matrigel (Clontech Laboratories, Inc., Mountainview, CA, USA) were put into a 24-well plate, thus forming upper and lower chambers. Cell suspension (200 µl) with density of 1×10⁵/ml was plated in the upper chamber and the medium of the suspended cells was free of FBS. The lower chamber was added with 400 µl RPMI-1640 supplemented with 20% FBS. After incubation for 24 h at 37°C, the cells which moved to the lower surface were stained with 0.5% crystal violet for 30 min and the cells on the upper chamber were removed. The cells in the lower chamber were observed with a microscope (BX51 Olympus; Olympus Corp., Tokyo, Japan) and counted at five random fields.

JAG1 was predicted to be a target gene of miR-26b by TargetScan online software (http://www.targetscan.org/vert_71/) and the two binding sites were located at 989–995 and 1,248-1,255 on JAG1 3′UTR. Genomic DNA acted as template to amplify JAG1 3′UTR sequences and then the 3′UTR sequences were cloned to pmirGlo plasmid (named pmirGlo-JAG1-WT). The binding sites were mutated and inserted in pmirGlo vector, named as pmirGlo-JAG1-MUT site1 and site2 (MUT S1 and MUT S2). In addition, the cells were co-transfected with the recombinant reporter plasmids (pmirGlo-JAG1-WT and pmirGlo-JAG1-WUT) and miR-211 mimic or its scramble negative controls using Lipofectamine 3000 (Invitrogen; Thermo Fisher Scientific, Inc.). After transfection for 48 h, luciferase activities were detected using Dual-Glo Luciferase Assay System (Promega Corp., Madison, WI, USA). Renilla luciferase activity was used for normalization.

miR-26b mimic and inhibitor were employed to overexpress or knockdown miR-26b (both from Guangzhou RiboBio Co., Ltd., Guangzhou, China). pcDNA3.1-JAG1 and its negative control (pcDNA3.1-NC) were designed and synthesized from Sangon Biotech Co., Ltd. (Shanghai, China). Human cervical cancer cells HeLa and JAR were seeded in 6-well plates and when the cells were 80% transfection was performed. pcDNA3.1-JAG1 and pcDNA3.1-NC were transfected with Lipofectamine 3000, whereas miR-26b mimic or inhibitor used Lipofectamine 2000 (both from Invitrogen; Thermo Fisher Scientific, Inc.).

The data were analyzed by Student's t-test and Pearson's χ² test using SPSS 19.0 software (IBM Corp., Armonk, NY, USA). One-way ANOVA, followed by Tukey's post hoc test, was employed to compare three or more groups. Survival was analyzed by Kaplan-Meier method with log rank test. The correlation between miR-26b and JAG1 was analyzed using Spearman's rank correlation coefficient. P<0.05 was considered to indicate a statistically significant difference.

The expression levels of miR-26b and JAG1 in cervical cancer tissues and cells were evaluated by RT-qPCR. miR-26b level was remarkably decreased in cervical cancer tissues compared with paracancerous tissues (P<0.001) (Fig. 1A). Moreover, in cervical cancer cells HeLa and JAR, miR-26b expression was lower than that in non-tumor epithelial Ect1/E6E7 cells (P=0.0026 and 0.0034, respectively) (Fig. 1B). The expression of JAG1, contrary to the expression of miR-26b, was overexpressed in cervical cancer HeLa (P=0.0008) and JAR (P=0.0012) cells versus Ect1/E6E7 (Fig. 1C). In addition, the expression of miR-26b and JAG1 were negatively correlated (P<0.0001, r=−0.5063) in cervical cancer tissues (Fig. 1D).

miR-26b mimic and inhibitor were transfected into cervical cancer HeLa (P=0.0026 and 0.040, respectively) and JAR cells (P=0.0038 and 0.0093, respectively) to overexpress or knock down miR-26b, which was measured by RT-qPCR (Fig. 2A). As predicted, in miR-26b mimic group cell migration and invasion were suppressed compared with negative control group in HeLa (P=0.0041 and 0.0015, respectively) and JAR (0.0233 and 0.0256, respectively) cells. Cell numbers were obviously increased in miR-26b inhibitor group compared with negative control group in HeLa (0.0144 and 0.0092) and JAR (0.0022 and 0.0003) cells (Fig. 2B). Thus, miR-26b suppresses cervical cancer cell migration and invasion.

TargetScan 4.2 (http://www.targetscan.org/vert_42/), a miRNA target identification tool, was utilized to search for potential target genes of miR-26b. We discovered that JAG1 is a potential target of miR-26b with two binding sites. The two binding sites were 5′-ACUUGAA-3′ from 989 to 995 and from 1,248 to 1,255 on 3′UTR (Fig. 3A). Two miR-26b potential wild-types (WT) of JAG1 3′UTR and corresponding mutant sites (MUT S1/S2; 5′-AGUACAU-3′ and 5′-AGCUCAA-3′) were constructed and co-transfected with miR-26b mimic or negative control into HeLa and JAR cells to confirm miR-26b binding to JAG1. Luciferase activities were reduced in HeLa and JAR cells when transfected with miR-26b mimic in MUT S1 (P=0.0038 and 0.0016, respectively), compared with negative control. However, the luciferase activity was not influenced by miR-26b mimic in MUT S2 (P=0.2307 and 0.6472), while it decreased in WT group (P=0.0003 and 0.0021), which reveals that miR-26b directly binds to site 2, rather than site 1 (Fig. 3B).

In addition, the mRNA (P=0.0132 and 0.0012) and protein level of JAG1 were reduced when overexpressed by miR-26b mimic in both HeLa and JAR cells. Also, miR-26b inhibitor could promote JAG1 expression (P=0.0006 and 0.0203) (Fig. 3C).

To further verify the miR-26b impact on cell migration and invasion via targeting JAR1, we detected JAG1 mRNA and protein level transfected with miR-26b and JAG1 in HeLa and JAR cells. As a result, when transfected with miR-26b mimic, the expression of JAG1 was reduced in both HeLa (P=0.0021) and JAR (P=0.0004) cells. When transfected with miR-26b mimic and JAG1, the mRNA and protein level of JAG1 (P=0.0425 and 0.0152) were increased, compared to transfection with only miR-26b mimic, which suggested that JAG1 could partially reverse the impact of miR-26b (Fig. 4A). When overexpressing miR-26b, the migration in HeLa and JAR cells (P=0.0043 and 0.0008, respectively) as well as invasion (P=0.0019 and 0.0022, respectively) were attenuated. Whereas, migration (P=0.0122 and 0.0232) and invasion (P=0.0251 and 0.0215) capacity increased when co-transfected with miR-26b mimic and JAG1, relative to transfection with miR-26b mimic alone (Fig. 4B). Thus, these results demonstrate that miR-26b inhibited cell migration and invasion by targeting JAG1 in cervical cancer.

Cervical cancer patients were segmented into different groups based on clinicopathological characteristics, in order to discover the relationship between miR-26b level and clinicopathological features of cervical cancer. It was found that miR-26b level was closely associated with FIGO stage (P=0.033), lymph node metastasis (P=0.025), serum SCC-Ag level (P=0.0327) and JAG1 level (P=0.030), while it had no association with age (P=0.860), tumor size (P=0.151), and histology degree (P=0.336) (Table I).

However, these clinicopathological factors can not be sufficient to accurately predict prognosis of patients. Patients were also set as miR-26b low expression group [miR-26b(−)] (n=28) if miR-26b expression was higher than the median value, and miR-26b high expression group [miR-26b(+)] (n=26) if miR-26b expression was higher than the median value. The overall survival (P=0.0259) and disease-free survival (P=0.0157) was obviously longer in miR-26b(+) group versus that of miR-26b(−) group (Fig. 4C).