| INTRODUCTIONGlioblastoma (GBM) is the most common primary malignant brain tumor in adults, with a median survival of approximately 14 months after diagnosis. As one of the standard therapeutic approaches, chemotherapy is effective to reduce tumor size, inhibit distant metastasis and prolong patient survival. However, GBM exhibits a high resistance to chemotherapy and recurrence is virtually assured. Therefore, it is increasingly challenging to understand the molecular mechanism of chemoresistance and develop effective therapeutic strategies to overcome drug resistance of GBM.Epithelial-mesenchymal transition (EMT) is a process during which cells undergo morphologic changes from epithelial phenotype to mesenchymal phenotype, resulting in enhanced motility and increased invasion, proliferation, and resistance to apoptosis. Tumor cells that have undergone EMT leave the primary tumor site, invade the extracellular matrix and basement membrane, colonize distant organs and form metastases. Emerging evidence indicates a strong link between resistance to chemotherapy and the induction of EMT in cancer. Therefore, determining the mechanisms that connect EMT and the development of drug resistance is essential for development of novel therapeutic strategies to overcome drug resistance.We examined the differential expression of a panel of genes between parental GBM cell line U87 and its counterpart drug-resistant cell line U87AR by cDNA microarray. The results showed that expression profiles of 21,325 genes changed significantly,1050 up-regulated genes including SNAI2 and 1120 down-regulated genes in U87AR cells as compared to U87 cells. We also found the expression of SNAI2 was significantly increased in relapsed GBM tissues than in the primary GBM specimens by immunohistochemistry.SNAI2 (also known as slug) belongs to the Snail superfamily of zinc finger transcription factors and has emerged as important repressor of E-cadherin and inducer of EMT. It has been reported that SNAI2 is involved in various biological processes, including cell migration and invasion, apoptosis, and angiogenesis. The expression of SNAI2 was positively correlated with the tumor grading. However, to our knowledge the role of SNAI2 in GBM chemoresistance has no documented.MicroRNAs (miRNAs) are small non-coding RNAs that participate in many cellular processes as essential gene regulators. They modulate protein expression by promoting RNA degradation and inhibiting transcription after binding to the 3’-untranslated region (3’-UTR) of mRNA. The differential expression profiles of miRNAs from normal tissue across cancers, and the surprising strength of these profiles play an important role in tumor classification and prediction of therapy response. Accumulating evidence has demonstrated that miRNAs have a key role in drug resistance and EMT. For example, miR-27a is reported to reverse cisplatin resistance on bladder cancer cells by targeting SLC7A11. MiR-135a modulates paclitaxel resistance by targeting APC in human non-small cell lung cancer cells. Moreover, recent study has shown that members of the miR-200 family play key roles in mediating the effects of TGF-β and other EMT regulators on EMT in breast cancer and lung adenocarcinoma cells. However, miRNAs and their target genes involved in EMT, resulting in resistance to chemotherapy, are still not fully understood.Our previous study showed that a number of miRNAs are involved in the drug resistance of GBM cells using miRNA microarray, we found that miR-203 is downregulation in U87AR cells than in the U87 cells by microarray and QRT-PCR. However, no data are available about miR-203 functions in GBM drug resistance. Besides, we analyzed the data from cDNA microarray in both U87 and U87AR cells. The gene expression profiling revealed that many mesenchymal marker genes (SNAI2, ZEB1 and vimentin) were upregulated in U87AR cells. Furthermore, by bioinformatics analysis, we found that the SNAI2 might be a potential target gene for miR-203. As mentioned above, we proposed the hypothesis that EMT involving the drug resistance of GBM may be under the regulation of SNAI2 and through EMT process.The purpose of this study was to explore the molecular mechanism that SNAI2 under the regulation of miR-203, through the induction of EMT, involved in regulation of drug resistance in GBM. Further enrich the molecular mechanism and provide a basis for the research of GBM drug resistance. Also, re-expression of miR-203 or targeting SNAI2 might serve as potential therapeutic approaches to overcome chemotherapy resistance in GBM.MATERIALS AND METHODSPatients and specimensPatient tissue samples were obtained from 999 Brain Hospital (Guangzhou, China), Zhujiang and Nanfang Hospitals (Southern Medical University, Guangzhou, China) between the period 2010/12 and 2013/10. Patients enrolled in this study included 9 grade â… astrocytoma cases,13 grade â…¡ astrocytoma cases,7 grade â…¢ astrocytoma cases, and 51 GBM cases. Among 51 GBM cases,16 GBM cases were relapsed 6 months after temozolomide therapy. Tissue specimens were snap-frozen in the operating room immediately during surgery and sent to a pathology department for diagnosis by a board-certified neuropathologist. For each patient, a frozen tumor sample (stored at-80 ℃) and a paraffin-embedded tissue specimen had to be available. Under the protocol approved by the Institutional Review Board, informed consents were obtained from the patients or their guardians according to institutional guidelines.Cell cultureHuman GBM cell lines U251 and U87, were obtained as a gift from College of Public Health of Southern Medical University, GuangZhou, China. The stable imatinib-resistant lines, U251AR and U87AR, were established and maintained in our laboratory. The cells were cultivated in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) containing 10% fetal bovine serum (Invitrogen), penicillin (200 units/ml) and streptomycin (100 μg/ml), and were incubated at 37 ℃ in a humidified incubator with an atmosphere of 5% CO2. To maintain the multidrug resistant phenotype, U251AR and U87AR were alternately fed with drug-free medium and medium containing 122 μg/ml of imatinib.Cell transfectionCells were transiently transfected with 100 nmol/1 of miR-203 mimics (miR-203), antagomirs (anti-miR-203) and miRNA negative control, or 60 nmol/1 short hairpin RNA (shRNA) specific to SNAI2, scrambled shRNA negative control (shNC) (Genepharma, Shanghai, China) by using Lipofectamine 2000 and OPTI-MEMI (Invitrogen). For stable transfection, SNAI2 expression plasmid (pcDNA3.1 SNAI2) and empty plasmid (pcDNA3.1 mock) were transfected into GBM cells by using Lipofectamine 2000. Positive transfectants were selected in 500μg/wl Geneticin (G418, Invitrogen). Individual colonies were harvested 24 h later for the evaluation of gene expression or functional assays. All of the RNA oligoribonucleotides were purchased from Genepharma (GenePharma, Shanghai, China) and the sequences are listed as following:RNA oligoribonucleotides SequencemiR-203 mimic 5’-GUGAAAUGUUUAGGACCACUAG-3’ 5’-AGUGGUCCUAAACAUUUCACUU-3’miR-203 mimic NC 5’-UUCUCCGAACGUGUCACGUTT-3’ 5’-ACGUGACACGUUCGGAGAATT-3’miR-203 inhibitor 5’-CUAGUGGUCCUAAACAUUUCAC-3’miR-203 inhibitor NC 5’-CAGUACUUUUGUGUAGUACAA-3’MiRNA microarray analysisMiRNA microarray analysis was done in the parental and resistant U87 lines. Briefly, total RNA was isolated using TRIzol (Invitrogen) and the RNeasy mini kit (Qiagen) according to the manufacturer’s instructions. The samples were labeled using the miRCURYTM Hy3TM/Hy5TM Power labeling kit (Exiqon) and hybridized on the miRCURY LNA Array (Exiqon, version 11.0). Scanning was performed with the Axon GenePix 4000B microarray scanner. GenePix pro version 6.0 was used to read the raw intensity of the image. Background subtraction and normalization were performed. We selected miRNAs whose expression levels between the parental and resistant U87 lines differed by at least 1.5-fold.RNA isolation, reverse transcription, and quantitative real-time PCRTotal RNA was extracted using Trizol reagent (Invitrogen) according to the manufacturer’s protocol. To quantitate miR-203 expression, total RNA was polyadenylated and reversely transcribed using miRNAs qPCR Quantitation Kit (Genepharma, Shanghai, China). To measure the mRNA levels of SNAI2, ZEB1, E-cadherin and vimentin, total RNA was reversely transcribed using primeScript RT reagent Kit (Takala, Dalian, China). Quantitative real-time PCR was carried out in ABI7500 sequence detection system (Applied Biosystems, Foster City, CA, USA) using SYBR Green according to the manufacturer’s instructions. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or U6 snRNA was used as an endogenous control. All samples were normalized to internal controls and fold changes were calculated through relative quantification (2-△△Ct). Sequences for primers were listed as following:Gene Primer sequencesSNAI2 Forward:5’-TGGTTGCTTCAAGGACACAT-3’ Reverse:5’-GTTGCAGTGAGGGCAAGAA-3’ZEB1 Forward:5’-GGGAGGAGCAGTGAAAGAGA-3’ Reverse:5’-TTTCTTGCCCTTCCTTTCTG-3’E-cadherin Forward:5’-CCCGGGACAACGTTTATTAC-3’ Reverse:5’-GCTGGCTCAAGTCAAAGTCC-3’Vimentin Forward:5’-CCCTCACCTGTGAAGTGGAT-3’ Reverse:5’-TCCAGCAGCTTCCTGTAGGT-3’GAPDH Forward:5’-GAGGTGATAGCATTGCTTTCG-3’ Reverse:5’-CAAGTCAGTGTACAGGTAAGC-3’miR-203 Order from GenePharma:Lot No.91126N22U6 Order from GenePharma:Lot No.8300871042Western blot analysisProtein lysates were separated by 10% SDS-PAGE, and electrophoretically transferred to PVDF (polyvinylidene difluoride) membrane (Millipore). Then, the membrane was incubated with rabbit monoclonal antibody against human SNAI2, ZEB1, E-cadherin and vimentin (Cell Signaling Technology, USA) followed by HRP (horseradish peroxidase)-labeled goat-anti rabbit IgG (Santa Cruz Biotechnology, USA) and detected by chemiluminescence. GAPDH was used as a protein-loading control.Luciferase reporter assayWe cloned the miR-203 response element (wide type or mutated) in the 3’-untranslated regions (3’-UTR) of SNAI2 into psiCheck2 plasmid downstream of luciferase reporter gene. Luciferase activities were measured using a luciferase assay kit (Promega, Madison, WI, USA), and target effect was expressed as relative luciferase activity of the reporter vector with target sequence over the one without target sequence.In vitro invasion assayCells growing in the log phase were treated with trypsin and re-suspended as single cell solutions. A total of 1×105 cells were seeded on a fibronectin-coated polycarbonate membrane insert in a transwell apparatus (Corning Inc., USA). In the lower chamber,600 μl DMEM with 10% FBS were added as a chemoattractant. After the cells were incubated for 48 h at 37 ℃ and 5% CO2 incubator, the insert was washed with PBS, and cells on top surface of the insert were removed by a cotton swab. Except the transwell membrane was precoated with ECMatrix and the cells were stained with crystal violet for 10 min and rinsed with PBS. The chambers were then photographed to compare the amount of invasive cells on the underside of the membrane for five predetermined fields (×200). All assays were independently repeated for at least three times.Wound healing assay and flow cytometric analysis of apoptosisCells were plated at a density of 5×105/cm2 in a six-well plate, and incubated at 37 ℃ in a CO2 incubator for 12 h, allowing cells to completely adhere and spread on the six-well plate. The cells grew and were maintained in serum-free DMEM for more than 24 h to create a confluent monolayer. The confluent monolayer was scraped with a sterile toothpick and washed with PBS. Serum-free DMEM was then added and the width of the wound gaps were measured using NIH Image J analysis and normalized to the time 0 wounds for four independent experiments. Apoptosis after transfection treatment was examined by using the Annexin V/Propidium Iodide Detection Kit (KeyGEN) according to the manufacturer’s instructions.In vitro drug sensitivity assay and cell viability analysisCells were plated in 96-well plates at 1×104 cells per well after transient transfection or adherence of stable transfected cells. After 24 h, the cells were treated with different concentrations of imatinib (Novartis, Basel, Switzerland), etoposide (VP-16) (Sigma Chemical Co., St. Louis, MO) and temozolomide (TMZ) (Sigma Chemical Co., St. Louis, MO), each at four concentrations ranging from 50 to 200 μg/ml for 48 h. The range of drug concentrations were based on earlier studies and aimed at obtaining IC50 values both for highly sensitive and resistant cases. The absorbance at 450 nm was measured after incubation with 10 μl of CCK-8 reagent (Dojindo, Molecular Technologies, Dojindo, Japan) for four hours. After shaking for one min, the spectrophotometric absorbance of the samples was determined by using Ultra Multifunctional Microplate Reader (Tecan) at 450 nm. The assay was conducted in five replicate wells for each sample and three parallel experiments were performed. For cell viability analysis, cells were plated in 96-well plates at 2×103 per well in a final volume of 100 μl. And cells were incubated with 50 μg/ml TMZ for 24,48,72,96 and 120 h. Cell growth were determined with CCK-8 according to the manufacturer’s instructions.Immunofluorescence stainingImmunofluorescence staining was done following the standard protocol. Briefly, cells were fixed with 4% paraformaldehyde, permeabilized in 0.5% Triton X-100, and blocked with 10% goat serum. The cells were then incubated overnight with specific primary antibody. After washing with PBS, the cells were incubated with fluorescence-conjugated secondary antibody for one hour. The slides were then washed with PBS and mounted with mounting medium containing Anti-fade reagent and 4’, 6-diamidino-2-phenylindole (DAPI). Cells were viewed under a fluorescence microscope.Immunohistochemical analysisFor immunohistochemical analyses of GBM, paraffin-embedded samples were sliced and mounted on microscopic slides. Rabbit monoclonal anti-SNAI2 and anti-E-cadherin antibodies (1:200 dilutions, Cell Signaling Technology, USA) were used as the primary antibodies. Heat-induced epitope was formed with a microwave in 10 mmol/1 citric acid buffer at pH 7.2. The samples were incubated with the antibody overnight in the same buffer followed by incubation with biotinylated secondary antibody (1:500 dilutions, Santa Cruz Biotechnology, USA). The bound antibodies were visualized by the avidin biotinylated peroxidase complex methods and diaminobenzidine tetrachloride (Santa Cruz Biotechnology).Statistical analysisAll statistical analyses were performed using SPSS 13.0 software (IBM Corporation, New York, NY, USA) and GraphPad Prism software 5.0 (GraphPad Software, Inc., San Diego, CA, USA). The results were presented as mean±s.d. of three replicate assays. Statistical analyses were performed using either an analysis of variance (ANOVA) or Student’s t-test. And statistical analysis of tissue specimens was performed using the Mann-Whitney test to evaluate the significance of differences between groups. The relationship between SNAI2 and miR-203 expression was explored by Pearson correlation. Kaplan-Meier survival curves were generated to evaluate the correlation of miR-203 expression levels with survival rate. All P<0.05 was considered to indicate statistical significance.RESULTSImatinib-resistant U251AR and U87AR cells exhibit EMT characteristicsBy using the parental cell lines U251 and U87, we previously established the imatinib-resistant GBM cell lines U251AR and U87AR, which had a cross-resistance to other anticancer drugs (etoposide/VP-16 and temozolomide/TMZ). We investigated whether the acquisition of the multidrug-resistant phenotype was accompanied by morphological changes of the cells. The parent U251 and U87 cells grew in clusters with tight cell-cell junctions, while U251AR and U87AR cells separated from one other and grew as loosely packed spindle-like cells. This suggested U251AR and U87AR cells had undergone EMT resulting in the acquisition of mesenchymal properties. We next investigated the expression and localization of a key epithelial marker (E-cadherin) in the imatinib-resistant GBM cells compared with their parental cells. E-cadherin predominantly localized at cell-cell contacts in U251 and U87 cells, while the staining intensity was reduced in U251AR and U87AR cells. Moreover, E-cadherin expression was significantly reduced at the mRNA and protein levels in U251AR and U87AR cells as compared with their parental cells. We also examined the expression of other EMT marker genes by performing gene expression profiling in both U87 and U87AR cells. The gene expression profiling revealed that a range of epithelial marker genes were downregulated and many mesenchymal marker genes were upregulated in U87AR cells. Additionally, qRT-PCR and western blotting analysis showed that mesenchymal genes ZEB1 and vimentin were also upregulated in U251 AR and U87AR cells.To examine whether EMT can promote cell invasion, we next performed a cell invasion assay which observed a significant increase in the invasive capacity of imatinib-resistant cells compared with their parental cells. Furthermore, cell viability assay showed that resistant GBM cells were significantly more capable of growth than their parental cells. All together, these data indicate that imatinib-resistant U251AR and U87AR cells have undergone EMT with enhanced invasiveness and increased cell viability.MiR-203 is downregulated in imatinib-resistant GBM cells and its re-expression sensitizes cells to anticancer drugs and reverses EMT-like propertiesTo screen miRNAs that are potentially involved in the acquisition of drug resistance and induction of EMT, we performed microarray miRNA analysis on U87AR and its parental U87 cells. Microarray analysis revealed a significant downregulation of 11 miRNAs and upregulation of 14 miRNAs in U87AR compared with U87 cells.MiR-203 was among the top downregulated miRNAs in U87AR cells and its downregulation was further validated by qRT-PCR. To explore the potential role of miR-203 in drug resistance and EMT, we used a miR-203 mimic (miR-203) and antagomir-203 (anti-miR-203) to modulate cellular levels of miR-203 in GBM cells. Expression of miR-203 was determined by qRT-PCR assay after miR-203 or anti-miR-203 was successfully transferred into U251AR or U87 cells, respectively.The half maximal inhibitory concentrations (IC50) values of anticancer drugs (imatinib, VP-16 and TMZ) in the imatinib-resistant cells and their parental cells transfected with miR-203 or anti-miR-203 were determined by cell counting kit-8 (CCK-8) assay to test the effect of miR-203 expression on the sensitivities of GBM cells to imatinib, VP-16 and TMZ. The IC50 values of imatinib, VP-16 and TMZ in the U251AR and U87AR cells transfected with miR-203 were significantly decreased by 1.9-3.3-fold, suggesting that upregulation of miR-203 expression could enhance the sensitivities of U251 AR and U87AR cells to all the three anticancer drugs. In contrast, the IC50 values of imatinib, VP-16 and TMZ in the U251 and U87 cells transfected with anti-miR-203 were increased by 2.4-3.2-fold, indicating that loss of miR-203 promotes resistance to anticancer drugs.Next, we asked whether miR-203 re-expression could reverse EMT-like properties of the imatinib-resistant GBM cells. MiR-203-transfected U251AR and U87AR cells showed epithelial cell features, characterized by cellular aggregation. Western blotting analysis showed that miR-203 significantly increased the expression of epithelial marker E-cadherin while decreased that of mesenchymal markers ZEB1 and vimentin in miR-203-transfected U251AR and U87AR cells. Furthermore, both transwell invasion and "wound healing" assays showed decreased invasion and migration activity of U251 AR and U87AR cells in the presence of miR-203, but remarkably enhanced invasion and migration activity upon anti-miR-203 treatment. Additionally, we found that miR-203 re-expression could induce cell apoptosis in U251AR and U87AR cells.Finally, to determine the potential clinicopathological implications of altered miR-203 expression, we examined the expression levels of miR-203 in the specimens from 80 patients with glioma by qRT-PCR. The relationships between clinicopathologic characteristics and miR-203 expression levels in patients with glioma are summarized. No significant association between miR-203 expression level and patient’s sex or age was observed in any of the 80 glioma cases. However, the expression of miR-203 was positively correlated with the tumor grading (WHO â… -â…¡ vs. WHO â…¢-â…£) (P=0.004, Mann-Whitney test) in glioma patients.SNAI2 is a direct target of miR-203We next explored the molecular mechanisms responsible for the drug resistance and EMT-suppressive effect of miR-203. The predicted target genes of miR-203 were retrieved from miRTarBase and TargetScan databases. SNAI2 (also known as slug), a transcriptional repressor of E-cadherin, was one of the potential candidates, which gained our attention, because of the importance of SNAI2 in EMT and drug resistance and the high conservation of the putative miR-203-binding sequences in the SNAI2 3’-UTR. Furthermore, we verified that SNAI2 mRNA was upregulated in U251AR and U87AR cells. To investigate whether SNAI2 is regulated by miR-203, we next transfected cells with miR-203 or anti-miR-203 and examined SNAI2 expression. qRT-PCR and western blotting analysis demonstrated that, in U251AR cells with high levels of SNAI2, restoration of miR-203 reduced the mRNA and protein expression of SNAI2. In contrast, in U87 cells, which express low levels of SNAI2, miR-203 inhibition by transfection with anti-miR-203 increased the mRNA and protein levels of SNAI2. These results demonstrate that miR-203 directly interacts with SNAI2 mRNA and represses its expression.To assess whether miR-203 directly regulates SNAI2 expression through the target site in the 3’-UTR of SNAI2, reporter constructs containing either the wild-type (WT) SNAI2 3’-UTR or SNAI2 3’-UTR with mutation at the predicted miR-203 target sequence were cotransfected into U251AR cells together with miR-203, miRNA control, anti-miR-203 or anti-miRNA control. Transduction of miR-203 caused marked inhibition of the WT SNAI2 3’-UTR, but had no effect on mutant SNAI2 3’-UTR. Meanwhile, miR-203 inhibition by anti-miR-203 substantially increased luciferase activities of WT SNAI2 3’-UTR compared with anti-miRNA control. All these results strongly suggest that SNAI2 is a direct target of miR-203 in GBM cells.Finally, to examine the pathological relevance of this interaction, we detected the expressions of miR-203 and its target SNAI2 in human GBM tissues. SNAI2 expression was negatively correlated with miR-203 expression in clinical GBM samples (Pearson’s correlation r=-0.402, p=0.003). In summary, SNAI2 is upregulated in the imatinib-resistant GBM cells and a direct target of miR-203, and their expression is negatively correlated in GBM patients.Loss of SNAI2 restores sensitivity to anticancer drugs and reduces the invasion and migration capacity of U251AR cellsThe direct targeting of SNAI2 by miR-203 led us to hypothesize that downregulation of SNAI2 by miR-203 in resistant GBM cells could be involved in drug resistance and/or EMT. For this purpose, we performed shRNA-mediated knockdown of SNAI2 in U251AR cells. The negative vector (shNC) and shSNAI2 were transfected into U251AR cells. At 48 hours post-transfection, fluorescent microscopy showed emission of green fluorescence. qRT-PCR and immunofluorescence analysis showed that SNAI2 expression level was significantly decreased in shSNAI2-transfected U251AR cells, compared to control cells. Knockdown of SNAI2 could indeed partially phenocopy the effects observed on sensitization to anticancer drugs upon overexpression of miR-203. Silencing of SNAI2 led to significant changes in cell morphology. The scattered, mesenchymal-like U251AR cells began to exhibit a more epithelial-like cobblestone appearance. The invasion and migration of U251 AR cells were markedly reduced after the inhibition of SNAI2 expression. Moreover, we examined the expression of those EMT marker that we had analyzed upon overexpression of miR-203 and also after knockdown of SNAI2. Silencing of SNAI2 increased E-cadherin expression and decreased the mesenchymal markers ZEB1 and vimentin, showing that the involvement of miR-203 in EMT could, at least in part, be via targeting SNAI2.SNAI2 contributes to chemoresistance and EMT in GBM cellsWe asked if overexpression of SNAI2 could induce drug resistance and EMT in parental GBM cells. For this purpose, we developed a stably SNAI2 over-expressing U87-pcDNA3.1-SNAI2 subline by transfection with pcDNA3.1-SNAI2. Overexpression of SNAI2 reduced the sensitivity of U87 cells to anticancer drugs and induced a shift in cell morphology from tight cell-cell junctions to loss of cell-to-cell contact. Also, enforced expression of SNAI2 promoted cell invasion in vitro. Furthermore, the downregulation of epithelial marker E-cadherin and the upregulation of mesenchymal markers ZEB1 and vimentin were observed in pcDNA3.1-SNAI2-transfected U87 cells.To further define the involvement of SNAI2 in the suppression of chemoresistance and EMT by miR-203, SNAI2 was transfected into miR-203-overexpressing U251AR cells. Next, we performed drug sensitivity assay to evaluate the chemoresistance changes in these cells. Ectopic expression of SNAI2 significantly rescued miR-203-induced inhibition of drug resistance. Moreover, reintroduction of SNAI2 markedly antagonized the inhibitory effect of miR-203 on cell invasion, abolished the mRNA expression of E-cadherin and restored ZEB1 and vimentin expression. These data suggest a crucial role of SNAI2 in driving chemoresistance and EMT of GBM cells.Low expression of miR-203 in GBM is associated with chemotherapeutic resistance and poor patient prognosisTo further evaluated the clinical significance of miR-203 expression in chemotherapeutic resistance and patient prognosis of GBM, SNAI2 expression was detected in tissues from 35 cases of patients with primary GBM and 16 cases of patients with relapsed GBM by immunohistochemistry. We found that the expression level of SNAI2 in relapsed GBM patients with treatment of temozolomide for 6 months was higher than that in primary GBM patients without treatment of temozolomide. In contrast, E-cadherin was lowly expressed in the relapsed GBM patients. Furthermore, qRT-PCR showed that the mRNA level of SNAI2 was significantly increased in relapsed GBM samples, whereas E-cadherin mRNA level was reduced compared to primary GBM tissues. Finally, we found that the expression of miR-203 was significantly reduced (p=0.0026) in relapsed GBM tissues. Moreover, patients with higher expression levels of miR-203 survived longer (p= 0.0017) than patients with lower expression levels.DISCUSSIONIn this study, we demonstrated that the imatinib-resistant U251AR and U87AR cells underwent EMT, and miR-203 was downregulated in these cells and clinical relapsed GBM specimens. Re-expression of miR-203 was capable not only of reversing EMT but also of sensitizing cells to anticancer drugs and reducing invasion and migration. Moreover, miR-203 suppressed the EMT and chemoresistance of GBM cells by targeting SNAI2. Our findings suggest that after developing drug resistance, miR-203 expression is reduced leading to a higher expression of SNAI2 and other targets, and the cells become more mesenchymal-like and invasive. These results are supported by clinical data where we found an inverse correlation between the expression of miR-203 and its target SNAI2 in GBM samples. Importantly, the significance and clinical relevance of miR-203 were further demonstrated in GBM patients.Consistent with our finding, increasing evidence demonstrates that miRNAs are associated with drug resistance and EMT in many types of tumors. Ujifuku et al. showed that miR-195, miR-455-3p and miR-10a upregulated in temozolomide (TMZ)-resistant GBM cells, played a critical role in acquired TMZ resistance. Similarly, downregulation of miR-181 was responsible for resistance to imatinib by directly targeting the Bcl-2 family member Mcl-1 in chronic myelogenous leukemia cells. Also, miR-221 and miR-222 were upregulated while miR-21, miR-342, and miR-489 were downregulated in tamoxifen-resistant MCF-7 cells; the reintroduction of miR-221 or miR-222 rendered the parent MCF-7 cells resistance to tamoxifen through inhibiting their target p27Kipl, which was reduced by 50% in resistant cells. However, not many reports explored the involvement of miRNAs in the development of multidrug resistance in cancer. Thus, it is a great strength of our present study which demonstrated that miR-203 reversed the resistance of GBM cells to three different anticancer drugs. Although studies have investigated the drug resistance role of a number of miRNAs in various tumors, the overwhelming majority of other miRNAs are yet to be further studied regarding the drug resistance.Numerous inducers of EMT in cancer cells have been identified including transforming growth factor-β, Wnt/β-catenin, Snail/SNAI2, Twist and talin. Recently, gain of Twistl and loss of E-cadherin observed in GBM specimens were correlated with shorter survival and poor temozolomide response in GBM patients, suggesting the involvement of EMT in GBM progression. Furthermore, acquisition of EMT features has been associated with drug resistance which could promote recurrence and metastasis after standard chemotherapeutic treatment. In this paper, we demonstrate that EMT also occurs in the imatinib-resistant GBM cells. Similarly, others characterized EMT in tamoxifen-resistant breast cancer cells. Ward et al. observed downregulation of epithelial marker (E-cadherin) expression that was accompanied by the upregulation of mesenchymal markers (TGFB2, SMAD3 and ZYX) and EMT transcription factor, SNAI2, in tamoxifen-resistant breast cancer cells compared with the parental cells. These findings support our own observations. Another report showed that EMT induction in anticancer drug-resistant cells was owing to increased levels of phosphorylated beta-catenin, which was associated with epidermal growth factor receptor (EGFR), leading to increased transcription of EMT regulators. However, in brain tumor, the role of miRNAs in co-regulation of EMT and drug resistance remained largely unexplored. This led us to investigate the potential role of miRNAs in EMT followed by acquired resistance to chemotherapy in GBM.Previous studies have shown that miR-203 is downregulated in metastatic prostate cancer and colorectal carcinoma. Furthermore, downregulation of miR-203 confers tumorigenicity ability to esophageal squamous cell carcinoma cells. A recent study by Ding et al. described the tumor suppressive effects of miRNA-203 in breast cancer, showing that loss of miRNA-203 led to increased invasion and metastatic potential of the cell system, similar to the results we have shown in GBM. However, other reports also suggest that elevated expression of miR-203 in pancreatic tumors is associated with poorer survival. We found that miR-203 suppressed EMT and chemoresistance by targeting SNAI2 in GBM cells. These results indicate that miR-203 may have a dual function as both a tumor suppressor and an oncogene, depending on the cellular context and tumor type. A number of groups have reported DNA methylation-mediated downregulation of miRNAs by proximal CpG islands. Thus, DNA methylation may be involved in the regulation of miR-203 in imatinib resistant GBM cells.Our present data further demonstrate that SNAI2 is a direct target of miR-203 and that miR-203-mediated inhibition of SNAI2 is dependent on a conversed motif in the 3’-UTR of SNAI2. Recent independent studies have shown that overexpression of SNAI2 alters cell invasion, motility, chemoresistance, metastasis and poor prognosis in several human cancers. As a member of the snail family of transcription factors, SNAI2 can repress E-cadherin transcription and induce EMT directly. Therefore, SNAI2 overexpression due to reduction of miR-203 may result in EMT and chemoresistance in GBM via these pathways. Additionally, miR-203 may relieve E-cadherin from transcriptional repression by targeting SNAI2 signaling. Nevertheless, because one single miRNA might have multiple targets, judicious considerations are essential for identification of the main functional targets.In summary, our study indicates that miR-203 is an effective inhibitor of EMT and ch... |
PDF Full Text Request