Introduction
Gliomas are the most malignant and prevalent brain tumors of glial origin and are divided into four clinical grades on the basis of their histology and prognosis. Grade IV GBM is the most malignant of all brain tumors. Despite recent advances in diagnostics and treatments, the prognosis of advanced patients suffering from GBM remains poor (1). One of the most important reasons for this dismal outcome is that the migration and invasion of GBM cells eventually lead to tumor recurrence and patient death. However, the factors that mediate GBM migration and invasion are still poorly understood.
MicroRNAs are a class of 20–25 nucleotide long non-coding RNAs that modulate gene expression post-transcriptionally by regulating mRNA translation or stability (2). The human miR-127 gene is located within a cluster on chromosome 14q32.31. The hsa-miR-127 precursor can express two mature miRNAs, miR-127-3p and miR-127-5p. Changes in the expression of miR-127 were previously observed in esophageal squamous cell carcinoma (3) and medullary thyroid carcinoma (4), and miR-127 was proposed to be a tumor suppressor in human bladder cancer cells (5). However, the role of miR-127 in GBM development has not been well documented and little is known in regards to its target genes.
We decided to investigate the roles of miR-127-3p in GBM. We found that miR-127-3p promoted cell migration and invasion using in vitro cell lines and in vivo mouse models. We further demonstrated that SEPT7 is a direct target of miR-127-3p and partially mediated the process of cell migration and invasion initiated by miR-127-3p. In addition, we performed microarray analysis after miR-127-3p overexpression was induced in a GBM cell line and found that the expression of many migration and invasion-related genes was significantly altered. Finally, immunofluorescent cell staining showed that miR-127-3p affected the remodeling of the actin cytoskeleton by regulating SEPT7 in GBM cells.
Materials and methods
Cell culture and reagents
Human GBM cell lines, LN229, T98G, A172, U87 and U251, were obtained from ATCC (http://www.atcc.org/) and grown in Dulbecco’s Modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Cells were incubated at 37°C and supplemented with 5% CO2. Antibodies specific to SEPT7 and GAPDH were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Horseradish peroxidase-coupled secondary antibodies were purchased from MultiSciences Biotech Co., Ltd. (Hangzhou, China). miR-127-3p inhibitors, which are antisense sequences of miR-127-3p with 2′-O-methyl modification for in vitro transfection, and their respective negative controls were obtained from Guangzhou RiboBio Co., Ltd. (Guangzhou, China).
Plasmid construction and establishment of stable miR-127-3p-overexpressing cells
The genomic sequence of the human miR-127-3p gene was PCR amplified using the primer 5′-GGAAGATCTGTAGTCCTGTCTGTTGGTCAG-3′ and 5′-CCCAAGCTTCCTGAAGAACTGCTTCCGCC-3′ from LN229 cells and cloned into pSUPER.neo (Oligo Engine, Seattle, WA, USA). It was designated as pSUPER.neo-miR-127-3p. Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was used for the transfection of the DNA plasmid according to the manufacturer’s protocol. Positive cells selected with 1 mg/ml G418 were confirmed by stem-loop RT-PCR. The full-length 3′UTR of SEPT7 was subcloned into the pmirGLO vector (Promega, Madison, WI, USA) to generate pmirGLO-SEPT7-3′UTR. Mutant construct of SEPT7-3′UTR, named pmirGLO-SEPT7-3′UTR-mut, which carried a substitution of 4 nucleotides within the core binding site of SEPT7-3′UTR, was carried out using Takara MutanBEST kit. The coding sequence of SEPT7 was synthesized and subcloned into the pUC57-simple vector (GenScript, Piscataway, NJ, USA), and then was amplified by PCR and subcloned into the pcDNA3.1 vector, named pcDNA3.1-SEPT7. The primers for SEPT7-3′UTR were 5′-CCGCTCGAGCTCTCTATTGACCACCAGTTAACG-3′ and 5′-AGCTTTCCTGCAGGCAGTGCCAATATATGGAAAATATC-3′. The primers for mutation were 5′-TAGGGTTTGACCAATTTGCACCAGTTTTATCC-3′ and 5′-CCAACAAACACTGATGTCCAAGCTGGC-3′. All PCR products were verified by DNA sequencing.
RNA quantification
Total RNA, containing miRNA, was extracted with TRIzol reagent (Invitrogen) following the manufacturer’s instructions. For mRNA analysis, reverse transcription was performed by using M-MLV reverse transcriptase (Promega). Amplification reactions were performed using the SYBR® Premix Ex Taq™ (Takara, Dalian, China). Data were normalized to the level of glyceraldehyde-3-phosphate dehydrogenase expression in each sample. Expression of miR-127-3p was quantified using stem-loop RT-PCR analysis as reported (6). The primers and other reagents for stem-loop RT-PCR were obtained from Guangzhou RiboBio. U6 RNA was used as an internal control. The 2−ΔΔCt method for relative quantification of gene expression was used to determine the expression levels of the transcripts.
Migration and invasion assays
For the migration assay, cells were plated onto 24-well Transwell chambers (Corning, Corning, NY, USA) with an 8-μm pore polycarbonate membrane. For the invasion assay, cells were plated on chambers precoated with ECM gel (Sigma, St. Louis, MO, USA). In both assays, the cells were plated in medium without serum, and medium containing 10% FBS in the lower chamber served as a chemoattractant. After 24 h, the cells on the upper surface were scraped and washed away, whereas the cells on the lower surface were fixed and stained with 0.05% crystal violet. Finally, six visual fields for each insert were randomly selected and counted under a microscope. The cell numbers were normalized to the control cell number, which was set to 100. The results were averaged from three independent experiments.
Animal model
Five-week-old female nude mice were purchased from Zhejiang Chinese Medical University (Zhejiang, China). All experimental protocols were approved by the Ethics Committee for Animal Experimentation of Zhejiang University. Exponentially growing LN229 cells (2×106) with stable ectopic expression of miR-127-3p or control vector were injected into nude mice through the tail vein (n=5). Eight weeks after injection, the animals were sacrificed. The liver and lung tissues were removed, sectioned, and stained with hematoxylin and eosin (H&E) stain to count the tumor loci by a pathologist.
Luciferase assay
LN229 cells were transiently transfected with the luciferase reporter gene constructs and miR-127-3p-expressing plasmid, and U251 cells were transiently transfected with the luciferase reporter gene constructs and miR-127-3p inhibitors. After 48 h, luciferase activities were measured using a dual luciferase reporter assay system according to the manufacturer’s protocol (Promega). For each plasmid construct, three independent transfection experiments were performed in triplicate.
Western blot analysis
The cells were lysed by a standard procedure in radioimmunoprecipitation assay buffer containing protease inhibitor cocktail (Calbiochem, Darmstadt, Germany). Protein concentrations of total cell lysates were measured using a BCA protein assay kit (Pierce Biotechnology, Rockford, IL, USA). Protein was then separated on polyacrylamide gels, transferred to a nitrocellulose membrane, incubated with the relevant antibodies, and detected with HRP-conjugated antibodies. Bands were visualized using ECL Western blotting detection reagents (Thermo Fisher Scientific, Rockford, IL, USA).
Microarray analysis
Two separate total RNA samples were extracted from LN229 cells with miR-127-3p overexpression and corresponding control cells. Genes regulated by miR-127-3p were identified by Affymetrix U133 Plus 2 GeneChip (Affymetrix, Santa Clara, CA, USA). The CEL files were imported into the Affymetrix’s Expression Console (V1.1) program, and the data were normalized using RMA normalization. Then the data were exported to the GeneSpring Program (Agilent, Inc., Palo Alto, CA, USA) for further analysis. In order to remove the low-expressed genes in the chips, probes with intensities <100 in both chips were excluded. Genes with absolute change of ≥2-fold and P<0.05 were considered differentially regulated by miR-127-3p. The online High-Throughput GoMiner program was used to analyze the Gene Ontology (GO) using biological processes terms at level 4. U251 GBM cells naturally present higher expression of miR-127-3p (as shown in Fig. 1), its array data (GSM803632) was downloaded from the GEO database on NCI 60 cell lines (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE32474). The same analysis pipeline was used to identify differentially expressed genes between LN229 with miR-127-3p and U251.
Immunofluorescence cell staining
Cells were seeded on coverslips and incubated for 24 h and then fixed with formalin for 15 min, washed with PBS, permeabilized for 3 min at room temperature with PBS containing 0.25% Triton X-100 and blocked for 30 min with PBS containing 1% of BSA and 0.5% Tween-20. Slides were incubated overnight at 4°C in a blocking solution containing anti-SEPT7 (Santa Cruz Biotechnology), washed with PBS, and then incubated in a blocking solution containing the secondary antibody conjugated with Cy3 (red) (Millipore, Darmstadt, Germany) and phalloidin-FITC (Beyotime Institute of Biotechnology, Haimen, China) for 1 h. After washing, coverslips were attached to glass slides. Cells were imaged using a fluorescence microscope.
Statistical analysis
Data are presented as means ± SD. Statistical comparisons of the studies were analyzed by Student’s t test or Chi-square test as appropriate. Correlations between groups were calculated with Pearson’s correlation analysis. A P value <0.05 was defined as indicative of a statistically significant result.
Results
miR-127-3p promotes human GBM cell migration and invasion
We initially analyzed the expression levels of miR-127-3p in five human GBM cell lines and found that there was great heterogeneity in miR-127-3p expression in the different GBM cell lines. Furthermore, we observed a statistically significant positive correlation between miR-127-3p levels and the cell migration or the cell invasion potential of the five GBM cell lines (r=0.89 and 0.93, respectively; P<0.05) (Fig. 1A). We observed a significant difference between the miR-127-3p high-expressing cell lines (U251 and U87) and low-expressing cell lines (LN229, A172 and T98 cells) in regards to migration and invasion potential (P=0.017 and 0.011, respectively; Student’s t-test).
To determine whether miR-127-3p is merely correlated or directly regulates human GBM cell migration and invasion, we selected LN229 and T98G cells, which are two cell lines with low endogenous miR-127-3p expression, and U251 cells, which have a high endogenous miR-127-3p expression, for further analyses. Fig. 1B and 1E show that we were able to induce miR-127-3p overexpression in the LN229 and T98G cells. We observed a >2-fold increase in cell migration and invasion potential in the LN229 and T98G cells with stable overexpression of miR-127-3p compared to the control cells (Fig. 1C, D, F and G).
In addition, we performed a complementary analysis by silencing miR-127-3p with an miR-127-3p inhibitor in U251 cells with high endogenous miR-127-3p expression (Fig. 1H). Inhibition of miR-127-3p in U251 cells led to ~40–50% reduction in migration and invasion compared with the control cells (Fig. 1I and J).
miR-127-3p increases tumor formation in a mouse model
To determine the effects of miR-127-3p on tumor invasive potential in vivo, LN229 cells either overexpressing miR-127-3p or transfected with a control vector were injected via the tail vein into nude mice. Eight weeks after injection, the animals were sacrificed. The liver and lung tissues were removed, sectioned, and stained with H&E to count the tumor loci by a pathologist.
In the mice inoculated with miR-127-3p-overexpressing cells, 5 of 5 cases had tumor formation in the liver and 2 of 5 cases had tumor formation in the lung. In contrast, 4 of 5 cases had tumor formation in the liver and none had tumor formation in the lung in the control group, both confirmed by pathohistological examinations (Fig. 2A and B). Although the number of tumors formed was higher in the miR-127-3p-overexpressing tumor group compared to the control cell group, the number was not statistically significant (P=0.29 and 0.11, respectively). However, as shown in Fig. 2C, the average number of tumors formed in the liver in the miR-127-3p-overexpressing group was 42, while it was 5 for the vector control group (P=0.02).
miR-127-3p targets the SEPT7 gene and its effects on cell migration and invasion are partially mediated through SEPT7
To explore the molecular mechanism of miR-127-3p function, we used the EIMMo program (http://www.mirz.unibas.ch/ElMMo2), which combined mRNA expression and targeted prediction algorithm, to predict targets of miR-127-3p in the human. We used the whole brain mRNA expression data set on the server to help predict miR-127-3p. One hundred and fifty-eight predicted targets were found. We found that SEPT7 had two entries (variant 1 and variant 2) that were among the top 10 predicted targets. Fig. 3A shows a miR-127-3p core binding site at nucleotides 67–74 of the SEPT7 3′UTR. In addition, the miR-127-3p target sequence at nucleotides 67–74 of the SEPT7 3′UTR is highly conserved among different species (Fig. 3A). We then decided to conduct additional analysis to determine whether SEPT7 is a target of miR-127-3p.
We constructed pmirGLO-SEPT7-3′UTR and pmirGLO-SEPT7-3′UTR-mut the latter contains mutated targeted sequences. Cotransfection of LN229 cells with pmirGLO-SEPT7-3′UTR and pSUPER.neo-miR-127-3p caused a 22% decrease in the luciferase activity compared with the negative control. Four-nucleotide substitution in the core binding site (pmirGLO-SEPT7-3′UTR-mut) abolished the suppressive effects (Fig. 3B). In addition, pmirGLO-SEPT7-3′UTR and pmirGLO-SEPT7-3′UTR-mut were cotransfected with either miR-127-3p inhibitor or the control oligonucleotide into U251 cells, expressing high levels of endogenous miR-127-3p. The inhibition of miR-127-3p by the inhibitors resulted in a significant increase in luciferase activity of the pmirGLO-SEPT7-3′UTR but not that of the pmirGLO-SEPT7-3′UTR-mut-transfected cells (Fig. 3C).
To test whether miR-127-3p expression affected endogenous SEPT7 expression, we transfected pSUPER.neo-miR-127-3p and the control plasmid into LN229 and T98G cells; a decrease in SEPT7 protein level in LN229 and T98G cells was observed (Fig. 4A, the left and the middle panel). Consistent with these results, silencing of miR-127-3p in U251 cells showed an increase in the SEPT7 protein level (Fig. 4A, the right panel). However, the SEPT7 mRNA level was not significantly influenced by overexpression or inhibition of miR-127-3p (Fig. 4B).
We then examined the SEPT7 protein levels in five GBM cell lines (Fig. 4C) and compared them with the miR-127-3p levels. In cell lines with high endogenous miR-127-3p (for example, U87 and U251; Fig. 1A, lanes 4 and 5), a low amount of SEPT7 protein at 49 kDa was observed (Fig. 4C, lanes 4 and 5), whereas cell lines with low miR-127-3p (for example, LN229, A172 and T98G; Fig. 1A, lanes 1–3) showed high amounts of SEPT7 protein (Fig. 4C, lanes 1–3). Across all five cell lines tested, we found a significant inverse correlation (P<0.05) between miR-127-3p and SEPT7 protein levels (Fig. 4D). We compared the quantified SEPT7 protein expression between the miR-127-3p high expression cell lines (U251 and U87) and low expression cell lines (LN229, A172 and T98 cells) and found that there existed a significant difference between the two groups (P=0.008; Student’s t-test).
We investigated whether EPT7 suppressed cell migration and invasion in LN229 and U251 GBM cells. As shown in Fig. 4E, overexpression of SEPT7 in U251 cells resulted in a decreased cell migration and invasion. To further establish a functional connection between miR-127-3p and SEPT7, we tested whether SEPT7 deregulation was required for miR-127-3p regulation of the migration and invasion of GBM cells. pcDNA3.1-SEPT7, which carried the whole SEPT7 coding sequence without 3′UTR and would thus not be suppressed by the miR-127-3p, and pSUPER.neo-miR-127-3p, were cotransfected into LN229 cells. We found that expression of SEPT7 partly abrogated the migration and invasion initiated by miR-127-3p in the LN229 cells (Fig. 4F).
miR-127-3p regulates genes related to migration and invasion
To identify the global effect of miR-127-3p on GBM cells, we performed microarray analysis to compare the gene expression profile of the miR-127-3p stably overexpressing LN229 cells with that of the LN229 control cells. Among 54,614 probe sets, we identified 1320 (791 downregulated and 529 upregulated) genes that were differentially expressed.
As the U251 GBM cells naturally present higher expression of miR-127-3p (Fig. 1), we compared its expression with LN229, and identified 5280 probe sets with 2-fold changes (data not shown). We then intersected the differentially expressed gene list for the comparison between LN229-miR-127-3p and LN229 mock control with that for the comparison between U251 and LN229 cells. We identified 855 probe sets that showed differential expression in both comparisons.
Using GO analysis we determined biological process gene categories that were significantly altered by miR-127-3p overexpression (FDR <0.01). Biological processes including cell proliferation, cell migration, cell motility, regulation of locomotion, which are fundamental biological processes required to achieve cell migration and invasion were found to be enriched (Table I).
miR-127-3p affects remodeling of the actin cytoskeleton by regulating SEPT7 in GBM cells
Dynamic regulation of the filamentous actin (F-actin) cytoskeleton is critical to cell adhesion and migration. We, therefore, examined changes in F-actin remodeling after miR-127-3p expression with a fluorescence microscope. In the control cells, the actin cytoskeleton formed a thick peripheral cytoskeleton near the cell membranes and the cells showed more rounded morphology (Fig. 5A, upper panel), while in the miR-127-3p-overexpressing LN229 GBM cells, the actin cytoskeleton formed thin stress fibers across the cytoplasm and the cells showed spindle-shaped morphology (Fig. 5A, lower panel). We then examined the F-actin remodeling of LN229 cells stably overexpressing miR-127-3p transfected with pcDNA3.1 or pcDNA3.1-SEPT7. The observed changes in cell morphology and the actin cytoskeleton structure in cells overexpressing miR-127-3p were rescued by SEPT7 overexpression (Fig. 5B).
Discussion
We investigated for the first time the role of miR-127-3p in GBM. Previous studies suggested that miR-127-3p might have tumor suppressing or promotive roles (3–5,7–10). Our inadvertent observation that there existed a statistically significant positive correlation between miR-127-3p expression and the migration and invasion potential of GBM cell lines prompted us to analyze the roles of miR-127-3p in GBM. We showed that miR-127-3p promoted migration and invasion of human GBM cells by in vitro cell analysis, and that miR-127-3p promoted metastasis in in vivo mouse models. This result is consistent with a demonstrated association of miR-127-3p with increases in lymph node metastasis in cervical carcinoma (11). This suggests that miR-127-3p may function as an oncomiR in GBM.
Accumulating evidence suggests that miRNAs play important roles in gliomas including regulation of their migration and invasion (12–14). As initial evidence for a role of miRNA in promoting glioma invasion, Gabriely et al (12) found that miR-21 promotes glioma invasion by targeting MMP inhibitors. Quintavalle et al (15) showed that miR-221/222 overexpession in human GBM increased its invasiveness by targeting the protein phosphate PTPμ. The inhibitory roles of miRNAs in glioma migration and invasion have also been reported. For example, miR-146b inhibits glioma cell migration and invasion by targeting MMPs (16). Ectopic expression of miR-26b, miR-124a and miR-34a in GBM cell lines resulted in significant inhibition of migration and invasion (13,17,18). miR-7 was downregulated in GBM and it was found to target insulin receptor substrate 2 (IRS2), epidermal growth factor receptor (EGFR) (19) and focal adhesion kinase (FAK). In addition, overexpression of miR-7 resulted in decreased migration and invasion, possibly mediated in part by decreased levels of MMP-2 and MMP-9 (14).
We found that SEPT7 was targeted by miR-127-3p. SEPT7 is a member of the septin family genes. Septins are a group of highly-conserved GTP binding proteins located at the plasma membranes in eukaryote cells (20), and 13 members of septins have been identified in the human (21). Septin members can act as either tumor suppressors or promoters. For example, the expression of ARTS (apoptosis-related protein in the TGF-β signaling pathway, the short isoform splice variant SEPT4_v2) is often lost in human leukemia (22). The loss of Sept4 function in mice was found to promote spontaneous leukemia or lymphoma (23). However, SEPT2 was found to be upregulated in human hepatoma carcinoma cells (HCC) and the phosphorylation of Ser218 of SEPT2 is crucial to the proliferation of human hepatoma carcinoma cells (24). SEPT1 plays an important role in spreading in squamous cell carcinoma DJM-1 cells (25). SEPT9 is overexpressed in diverse human tumors including breast, cenral nervous system (CNS), endometrial, kidney, liver, lung, lymphoid, oesophageal, ovarian and pancreatic cancer (26).
The SEPT7 gene is located on human chromosome 7p14.3, and alternative splicing of the gene results in the transcription of three isoforms. Although SEPT7 is expressed in most tissues, it is especially abundantly expressed in brain tissues (26). SEPT7 is involved in several human diseases including neoplasia (27,28) and neurodegenerative disorders (29). Previous data suggest that SEPT7 is a tumor suppressor. SEPT7 was reported to be downregulated in low-grade diffuse astrocytomas (30) and in other CNS tumor tissues compared with normal tissues (26,27). In addition, the high level of SEPT7 expression in neuroblastoma and glioma may be associated with favorable clinical and biological characteristics (27,28). SEPT7 has been shown to inhibit the migration and invasion of GBM cell lines (31). This observation indicates that SEPT7 might have a negative role in migration and invasion of GBM cells. Our observation in GBM U251 cells confirmed the suppressive role of SEPT7 on cell migration and invasion (Fig. 4E).
To establish the functional connection between miR-127-3p and SEPT7, we overexpressed SEPT7 and miR-127-3p in LN229 cells and found that SEPT7 protein could partly abrogate the function of miR-127-3p (Fig. 4F). Therefore, SEPT7 protein might be a mediator of the regulation of cell migration and invasion by miR-127-3p. Nonetheless, overexpression of SEPT7 did not completely abolish the effect of miR-127-3p, suggesting that additional mediators might exist. Additional experimentation is necessary to identify these mediators.
Our expression profile analysis of miR-127-3p-regulated genes provided a mechanistic link with its role in promoting migration and invasion as we found that biological processes such as cell migration, cell motility, regulation of locomotion are enriched in miR-127-3p-regulated genes (Table I).
Our data demonstrated that cells with high levels of miR-127-3p exhibited a spindle-shaped morphology and the actin cytoskeleton formed thin stress fibers across the cytoplasm, in contrast with more rounded morphology and a thick peripheral actin cytoskeleton near cell membranes shown in the control cells (Fig. 5). Our data is consistent with reported roles of SEPT7 in cytoskeleton remodeling, cell shape and cell movement (32–35). Recently, SEPT7 was reported to affect the actin cytoskeleton and cellular morphology via the SOCS7/NCK signaling pathway (29). There is also evidence showing that upregulation of SEPT7 inhibits invasion of GBM cells by distribution of α-tubulin (31). Collectively, these data suggest that miR-127-3p regulates cell morphology and the structure of the actin cytoskeleton by targeting SEPT7. Our observation that reintroduction of SEPT7 into LN229 cells overexpressing miR-127-3p rescues the phenotype seems to confirm this hypothesis.
In conclusion, our findings revealed that miR-127-3p promoted the migration and invasion of GBM cells partly by targeting SEPT7. miR-127-3p may be explored as a novel biomarker for predicting the migratory and invasive potential of glioma cells or may be exploited as a therapeutic target for GBM.