loss of mir-204 expression enhances glioma - cancer research

Tumor and Stem Cell Biology

Loss of miR-204 Expression Enhances GliomaMigration andStem Cell-like Phenotype

Zhe Ying1,3, Yun Li1,3, JuehengWu1,3, Xun Zhu1,3, Yi Yang2,3, Han Tian1,3, Wei Li1,3, Bo Hu4, Shi-Yuan Cheng4,and Mengfeng Li1,3

AbstractPhenotypic similarities have long been recognized between subpopulations of glioma and neural stem cells.

Many of these similar properties, including the robust abilities to self-renew,migrate, and invade, are hallmarks ofglioma cells that render them extremely aggressive. However, the molecular mechanisms underlying thischaracter, particularly in glioma stem-like cells that drive this disease, remain poorly understood. Here, wereport the results of a differentialmiRNA expression screen that compared glioma and neural stem cells, wherewefound that miR-204 was markedly downregulated in both types of cells. Mechanistic investigations revealed thatmiR-204 simultaneously suppressed self-renewal, stem cell-associated phenotype, and migration of glioma cellsby targeting the stemness-governing transcriptional factor SOX4 and the migration-promoting receptor EphB2.Restoring miR-204 expression in glioma cells suppressed tumorigenesis and invasiveness in vivo and increasedoverall host survival. Further evaluation revealed that the miR-204 promoter was hypermethylated and thatattenuating promoter methylation was sufficient to upregulate miR-204 in glioma cells. Together, our findingsreveal miR-204 as a pivotal regulator of the development of stem cell-like phenotypes and cell motility inmalignant glioma cells. Cancer Res; 73(2); 1–10. �2012 AACR.

IntroductionGliomas are the most common malignant primary brain

tumors in adults and exhibit a spectrum of aberrantly aggres-sive phenotype. Despite advances in treatments during pastdecades, the prognosis of the disease remains poor, with themedian survival time approximately 12 to 14 months. Theextremely poor prognosis of patients with gliomas is largelydue to the high tendency of tumor invasiveness, which leads tosevere structural and functional damage to the surroundingbrain tissue, incomplete surgical resection, and high frequencyof tumor recurrence (1). The molecular mechanisms underly-ing the aggressive malignant phenotype of glioma cells, how-ever, remain largely unknown.

It has been established that glioma cells, or their subpopula-tions, share similar biological phenotype and gene expressionprofiles to those of neural stem cells (2, 3). Two commonphenotypic features shared by glioma cells and neural stemcells are their robust abilities to migrate through brain paren-chyma (4, 5) and self-renewal (6, 7). For neural stem cells, thesecharacters typical of cellular "stemness" are essential for tissuehomeostasis, regeneration, and repairing processes in thecentral nervous system. In malignant gliomas, however, theseproperties contribute to tumor invasiveness. Understandingwhether there are common factors that simultaneously mod-ulate the malignant phenotypes of glioma cells and the stem-ness of neural stem cells is expected to provide new insights indeveloping novel and effective therapies for glioma. In thiscontext, identification of genetic and/or epigenetic factors thatmodulate the "stemness" phenotype of glioma cells is of greatimportance.

Deregulation of miRNAs has been implicated in the devel-opment and progression of nearly all tumor types (8). miRNAsplay pivotal roles in development, particularly in modulatingstem cell-specific pathways (9). This prompted us to identifyand study miRNAs with common or similar expression pat-terns in neural stem cells and glioma cells. In this report, wedescribe the miR-204 as a miRNA that is significantly down-regulated in both gliomas and neuronal stem cells, throughmethylation of its promoter. Reintroduction of miR-204 intoglioma cells markedly suppresses the motility and invasion ofglioma cells as well as their stem cell-like phenotype throughtargetingmultiple regulators. Furthermore, our data show thatmiR-204 inhibits glioma tumorigenesis and invasion in thebrain of mice. These findings, thus, reveal a novel pathway by

Authors' Affiliations: Departments of 1Microbiology and 2Pharmacology,ZhongshanSchool ofMedicine, SunYat-senUniversity; 3KeyLaboratory ofTropical Disease Control (Sun Yat-sen University), Chinese Ministry ofEducation, Guangzhou, Guangdong, China; and 4Department of Neurol-ogy, Northwestern Brain Tumor Institute, Center for Genetic Medicine &The Robert H. Lurie Comprehensive Cancer Center, Northwestern Univer-sity Feinberg School of Medicine, Chicago, Illinois

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Z. Ying and Y. Li contributed equally to this work.

Corresponding Author:Mengfeng Li, Sun Yat-sen University, ZhongshanSchool of Medicine, 74 Zhongshan Road II, Guangzhou, Guangdong510080, China. Phone: 86-20-87332748; Fax: 86-20-87331209; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-12-2895

�2012 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org OF1

Research. on December 25, 2018. © 2012 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 29, 2012; DOI: 10.1158/0008-5472.CAN-12-2895

http://cancerres.aacrjournals.org/

which epigenetic modulation of miR-204 renders the abilitiesof robust self-renewal and aggressive invasion of glioma cells.

Materials and MethodsCell lines and primary cell culture

Primary normal human astrocytes (NHA) were purchasedfrom Sciencell Research Laboratories (San Diego, CA) andcultured under conditions suggested by the manufacturer.Glioma cell lines, including LN443, LN444, LN464, U118MG,T98G, U251MG, U87MG, D247MG, A172, LN319, LN382T,LN229, LNZ-308, and SNB19, were shared with Dr. Shi-YuanCheng's laboratory at Northwestern University, Chicago, Illi-nois and cultured in Dulbecco's Modified Eagle's Medium(DMEM) supplemented with 10% FBS (HyClone) as previouslydescribed (10, 11). All cell lines were authenticated by shorttandem repeat fingerprinting at IDEXX RADIL (Columbia,Montana) and Services at the SYSU Forensic Medicine Lab(Guang Zhou China).

Tissue specimens and patient informationTwenty freshly dissected glioma tumor specimens were

clinically and histopathologically diagnosed at the First Affil-iated Hospital of the Sun Yat-sen University. Six normal braintissueswere obtained bydonation from individualswhodied intraffic accidents and confirmed to be free of any pathologicallydetectable lesions. Prior consents of donors and approvalsfrom the Institutional Research Ethics Committee wereobtained.

Plasmid, retroviral infection, and transfectionA DNA fragment containing the hsa-miR-204 precursor

flanked by a 500-bp genomic sequence at either end wasinserted into the retroviral vector pMSCV-puro (Clontech).The open reading frames (ORF) of SOX4 and EphB2 genesgenerated by PCR were cloned into the retroviral vectorpMSCV-neo (Clontech). Retroviral production and infectionwere carried out as previously described (10). 30-Untranslatedregions (UTR) of SOX4 and EphB2 were amplified and clonedinto the downstream region of a luciferase gene in a modifiedpGL3 control vector (Promega). ThemiR-204mimics andmiR-204 inhibitor oligonucleotides and their corresponding controloligonucleotides were synthesized at Ribo Biotech (Guangz-hou, China). Transfection of plasmids or oligonucleotides wascarried out using the Lipofectamine 2000 reagent (Invitrogen)according to the manufacturer's instruction.

Single cell tracks assayCells (5 � 103/well) were seeded in 6-well plates. Single cell

tracks were obtained by the Axio Observer-Z1 time-lapsemicroscopy system (Carl Zeiss). Experiments were conductedaccording to the manufacturer's instruction and repeatedindependently at least 3 times.

Flow cytometric analysisCells were dissociated into single-cell populations and

labeled with a phycoerythrin-conjugated CD133/2 (293C3)antibody (Miltenyi Biotec). The expression level was analyzed

using the EPICS XL flow cytometer with EXPO32 ADC software(Beckman Coulter).

Tumor xenograft live imagingGlioma xenografts stably expressing firefly luciferase togeth-

er with miR-204 or the corresponding control vector wereorthotopically implanted in the brains of nude mice as previ-ously described (11) and monitored using the IVIS SpectrumLive Imaging System (Caliper Life Sciences). Image calibrationand visualization were carried out using the Life Imagine 4.2software (Caliper Life Sciences).

Bisulfite genomic sequencingGenomic DNA from primary NHA, glioma cell lines, and

clinical specimens was bisulfite-modified with the EpitectBisulfite Kit (Qiagen). Bisulfite-treated DNA was amplifiedwith bisulfite-sequencing PCR (BSP) primers located in theTRPM3 promoter. PCR products were cloned using the pGEM-T Easy Vector System (Promega). Plasmids from single colo-nies were purified and sequenced.

Microarray data visualizationTCGA dataset were obtained from (12). GEO expression

profile was obtained from (13). All microarray data wereprocessed on the MeV 4.6 platform (14).

Statistical analysisStatistical analyses, with the exception of the microarray

data, were carried out using the SPSS 11.0 statistical softwarepackage. P values of less than 0.05 in all cases were consideredstatistically significant.

ResultsmiR-204 was downregulated in gliomas

ToidentifymiRNAswithsimilar expressionpatterns ingliomaand neural stem cells, we retrieved miRNA expression profilesfrom2 datasets, The Cancer GenomeAtlas (TCGA) and theGEOdataset GSE29759. As shown in Fig. 1A, miR-204 emerged as oneof the most prominent downregulated miRNAs in the TCGAglioblastoma (GBM) dataset, accompanied by several otherdownregulated miRNAs reported previously, namely, miR-124,106a,17-5p and members of the miR-181 family (15, 16). Asdocumentedpreviously,miRNAsupregulated ingliomas, includ-ing miR-21 (17–19), miR-9 (20, 21), and miR-10b, were alsoscreened out. (22, 23). Comparing the miRNA profiles in theTCGAdatasetwiththose inneural stemcells identifiedthatmiR-204 was downregulated in both glioma and neural stem cells(Fig. 1B). To validate the expression pattern of miR-204 ingliomas, quantitative reverse transcription PCR was conductedwith 6 normal brain tissue samples and 20 freshly dissectedglioma samples. Consistentwith thedatapresented in theTCGAprofile, expression of miR-204 was markedly downregulated inhigh-grade glioma samples (WHO tumor grades III and IV) andto a lesser degree, decreased in WHO tumor grades I and IIglioma samples, compared with normal brain tissues (Fig. 1C).Notably, 10 out of 14 glioma cell lines displayed significantlydownregulated expression of miR-204 relative to that in 2primary cultured normal human astrocytes (NHA; Fig. 1D).

Ying et al.

Cancer Res; 73(2) January 15, 2013 Cancer ResearchOF2




Restoration of miR-204 in glioma cells suppressedcellular migration and invasionTo investigate whether miR-204 could modulate the migra-

tion and invasiveness of glioma cells, the SNB19 and LN382Tglioma cell lines, which possessed the lowest levels of miR-204expression among all tested glioma cell lines, were selected forfurther studies. By retrovirally transducing glioma cells withmiR-204 at expression levels comparable with that of NHAs(Supplementary Fig. S1), we first examined the effect of miR-204 on cell migration using the wound-healing assay. As shownin Fig. 2A, compared with the vector-control cells, whichspread to the center line within 16 hours, miR–204-transducedcells exhibited considerably slower migration and decreasedcell spreading. Furthermore, we tracked the movement ofglioma cells, using a time-lapse imaging system, and examinedwhether miR-204 altered the migrating behavior of individualglioma cells. As shown in Fig. 2B, miR–204-transduced-SNB19and -LN382T cells generally retained their original positions,whereas vector-control cells moved away to a higher extent.Specifically, wemeasured the migration tracks of 10 individualcells within 5 hours and observed that miR–204-transducedcells moved within a very limited area. In contrast, the controlcells spread out in a markedly larger area, with approximately1.5- to 3-fold of increase in migration distance (Fig. 2C). Tofurther determine the effect of miR-204 expression on motility

of glioma cells, a transwell penetration assaywas conducted. Asshown in Fig. 2D, less number of miR-204 cells invaded acrossthe membrane precoated with Matrigels when compared withcontrol cells. Importantly, the 3-dimensional (3D) spheroidinvasion assay showed that vector-control cells displayed ahighly aggressive penetrating growth after 5 days, whereas themiR-204 transduced-cells grew in spherical colonies (Fig. 2E).Taken together, these observations suggested that miR-204strongly inhibited the migration and invasion of glioma cells.

miR-204 Attenuated stem cell-like phenotype of gliomacells

Malignant gliomas tend to relapse after surgical resection,and this characteristic is believed to be largely attributable tothe stem cell-like properties of a fraction of cells within aglioma tumor (2). As miR-204 is expressed at very low levels inboth glioma and neural stem cells but at high levels indifferentiated normal brain tissues, we determined the poten-tial role of mir-R-204 in the development and maintenance ofthe stem cell-like property of glioma cells. Sphere-formingassay showed (Fig. 3A) that miR-204 overexpressing-cellsformed much smaller neural spheres after 7 days of culturewhen compared with control cells (�2-fold smaller in diam-eter), indicating markedly decreased self-renewal ability bymiR-204.

Figure 1. miR-204 expression is downregulated in glioma andneuronal stemcells. A, expression profiles ofmiRNAwere obtained from the TCGAdatabase anddifferential expressed miRNAs were retrieved. B, most differentially expressed miRNA between neural stem cells (NSC) and differentiated counterparts(Dif-NSC)were obtained frompublished expression dataset GSE29759 of theNCBIGEOdatabase. C, qPCR analysis ofmiR-204 expression in 6 normal brainspecimens without identifiable pathological lesions, 6 WHO tumor grades I and II, and 14 WHO tumor grades III and IV glioma specimens. D, relativemiR-204 expression in indicated primary cultured astrocytes (NHA, A andB) andglioma cell lineswas obtained byqPCR.Error bars representmean�SD from3 independent experiments. ��, P < 0.001.

miR-204 Suppresses Glioma Stem Cell-like Phenotype

www.aacrjournals.org Cancer Res; 73(2) January 15, 2013 OF3




To further investigate the potential role of miR-204 insuppressing the stemness property of glioma cells, we exam-ined the expression of pluripotency factors in neuronal spheresformed by miR–204-transduced cells and control cells. Asshown in Fig. 3B, the expression levels and patterns of plur-ipotency factors, including SOX2, NANOG, KLF4, and OCT4,were similar between spheres formed by vector control- andmiR–204-transduced cells. However, at 48 hours after stim-ulation with 1% serum, miR–204-expressing cells rapidly losttheir expression of pluripotency factors, retaining only 2% ofexpression in the neuronal spheres after induction for dif-ferentiation, while the vector-control cells retained expres-sion of pluripotency factors at approximately 30% of the

initial levels (Fig. 3B). Consistent with these results, at bothtime points, CD133-positive proportions were significantlylower in miR–204-transduced cells than those in control cells,and a faster declining rate of CD133 positivity was alsodisplayed in miR-204-transduced cells (Fig. 3C). In furtherexperiments, we co-stained for a glioma stem cell marker,Nestin, and an astrocyte differentiation marker, GFAP, duringthe process of differentiation. We observed a significantlyfaster decrease of Nestin and simultaneous increase of GFAPexpression in cells transduced with miR-204 as comparedwith those in control cells (Fig. 3D). Taken together, thesedata showed that reintroduction of miR-204 in glioma cellsreduced the stem cell-like population in glioma and greatly

Figure 2. miR-204 suppresses gliomamigration and invasion in vitro. A, wound-healing assay was conducted with indicated cells and images were taken at 0,8, and 16 hours. B, growth pattern of indicated cells were monitored using time-lapse microscopy. Images comparatively show the pattern of cell growth at 0and 24 hours. C, 10 representative cell migration tracks were obtained from 5 hours of time-lapse image of indicated cells, and their initial positions werealigned with the origin of coordinates (left). Cumulated migration distance of indicated cells in 5 hours was measured. D, representative images (left) andquantification (right) of penetrated cells were analyzed using the Transwell matrix penetration assay. E, representativemicrographs of indicated cultured cellsat day 7 of culture in 3-dimensional spheroid invasion assay. Error bars represent mean � SD from 3 independent experiments; �, P < 0.05; ��, P < 0.01.

Ying et al.





attenuated the ability of stem cell-like glioma cells to retainstemness.

Reintroduction of miR-204 induced loss of invasion andtumorigenecity in vivoNext, we test whether miR-204 is involved in tumorigenesis

and invasiveness of glioma cells in vivo. An orthotopic gliomaxenograft model was employed by xenotransplanting lucifer-ase-expressing glioma cells transduced with miR-204 or cor-responding control vectors in the brains of nude mice. Tumorgrowth was monitored using a live animal bioluminescenceimaging system during the experiment. As shown in Fig. 4A,both miR–204-transduced cell lines exhibited significantlyslower growth after implantation. For the highly proliferativeSNB19 cells, results from luminescence images revealed a highdegree of suppression on the growth of miR–204-transduced

cells in contrast to the larger intracranial tumor formed by thecontrol cells at day 10 after implantation. All mice bearingvector-control cells died 18 days after implantation, while only1mouse implanted withmiR–204-transduced SNB19 cells diedat day 21 (Fig. 4B). Similar inhibition of tumorigenicity of miR-204 transduction was also observed in LN382T cells, which areknown to growmore slowly. Specifically, 1mousewithmiR-204cells died after 35 days of tumor cells implantation, and allvector-cell-implantedmice were dead at day 29 (Fig. 4B). Thesedata suggested that miR-204 was a strong inhibitor of gliomatumorigenicity in vivo.

Furthermore, immunohistochemical (IHC) analysis showedthat tumors formed by vector-control cells exhibited an exten-sive branch-like growing pattern that spread into the sur-rounding tissue. By contrast, miR–204-overexpressing cellsformed an oval-shaped tumor with smooth margins and

Figure 3. miR-204 promotesdifferentiation of stem cell-likepopulation of glioma cells. A, top,representative images of neuralspheres formed by indicated cellsafter 7 days of culture. A, bottom,statistical analysis of the averagediameter of spheres. B, qPCRanalysis of expression ofpluripotency factors, includingSOX4, NANOG, KLF4, and OCT4, atindicated time points after inductionof differentiation in indicated spherediverted cells.C, expression of CD-133 wasmeasured by FACS in indicatedspheres-derived cells 24 and 48hours after induction ofdifferentiation. D, co-staining ofGFAP and Nestin in indicatedspheres-derived cells 24 and 48hours after induction ofdifferentiation. Error bars representmean � SD from 3 independentexperiments; �, P < 0.05;��, P < 0.01.






noninvasive front (Fig. 4C), further validating that the invasivebehavior of glioma could be suppressed by miR-204 in vivo.

Core regulators of migration and stemness were directtargets of miR-204

As miRNAs exert their functions by targeting multipletranscripts that are coordinately orchestrated in a biologicalprocess, we screened for targets of miR-204 using the TargetS-can Program and identified 2 conserved binding sites for miR-204 in the 30UTR region of the SOX4 gene, a known coreregulator governing the stemness of both glioma and neuralstem cells (24, 25). Meanwhile, EphB2, which is known to berequired for themigration of both glioma and neural stem cells(26–29), was also found as a potential target gene of miR-204(Fig. 5A). To validate whether SOX4 and EphB2 were targets ofmiR-204, we transiently expressed miR-204 in SNB19 andLN382T cells, both of which possess low endogenous miR-204 expression. Significantly decreased protein expression ofboth SOX4 andEphB2was observed after transfection of amiR-204 mimic oligonucleotide. In contrast, inhibition of miR-204by transfecting a miR-204 inhibitor oligonucleotide in LN443and U251MG cells, which express high levels of endogenousmiR-204, led to elevated expression of both proteins (Fig. 5B).Furthermore, in the in vivo glioma model described earlier,lower levels of SOX4 and EphB2 staining were detected intumors formed bymiR-204-transduced cells, as comparedwiththose in the control tumors (Supplementary Fig. S2).

To determine whether inhibition of miR-204 on SOX4 andEphB2 expression ismediated by posttranscriptional effects onthe 30-UTRs of both genes, we constructed a luciferase reporter

plasmid containing 30-UTRs for SOX4 and EphB2, respectively.As shown in Fig. 5C, ectopic expression of miR-204 markedlydecreased, while inhibition of miR-204 was significantly ele-vated with the reporter luciferase activities (Fig. 5C), suggest-ing that miR-204 posttranscriptionally inhibited SOX4 andEphB2 expression. Furthermore, mutagenesis in miR-204 tar-get sites of the SOX4 or EphB2 30-UTR linked to the luciferasereporter confirmed the site-specific effect of miR-204 (Fig. 5C).To examine the interaction between miR–204-containingmiRNP and target mRNAs, we pulled down Ago1 and itsassociated RNA in cells transfected by control miRNA or themiR-204 mimic. Quantitative PCR (qPCR) assessment of theSOX4 and EphB2 30-UTR regions revealed high levels of SOX4and EphB2 transcripts enrichment in the miRNP of miR-204mimic-transfected cells, strongly suggesting that miR-204could guidemiRNP to target SOX4 and EphB2 (Fig. 5D). Finally,we assessed the expression of SOX4 and EphB2 in 20 gliomaspecimens shown in Fig. 1C, and found significant inversecorrelation between miR-204 and its 2 targets (SupplementaryFig. S3). This observation indicates miR-204 may act as aregulator for SOX4 and EphB2 expression in clinical gliomaspecimens.

To determine the functional significance of SOX4 and EphB2genes in miR–204-induced phenotype, we further stablyexpressed SOX4 ORF, or EphB2 ORF, that did not contain the30-UTR in miR–204-transduced cells. As shown in Fig. 5E,restored SOX4 expression could significantly abrogate theinhibitory effect of miR–204-induced sphere formation. Sim-ilarly, reexpression of EphB2 ORF in miR–204-transduced cellsrestored the invasiveness of glioma cells induced by miR-204

Figure4. ReintroductionofmiR-204suppresses glioma tumorigenesisand invasiveness in vivo. A, micewith intracranial glioma xenograftswere monitored by luciferase liveimaging system at indicated timepoints. Heat-map scale barrepresents photon emission. B,survival curves of mice with brainglioma xenograft formed byindicated cells. C, H&E staining oftumor boundary of gliomaxenograft formed by indicatedcells.

Ying et al.





expression (Fig. 5F). Taken together, these data revealed thatby coordinately suppressing core regulators of glioma stem-ness and migration, miR-204 conferred the abilities of self-renewal and invasion in glioma cells.

Hypermethylation in the promoter region of themiR-204geneWe next sought to investigate the molecular mechanism

that mediates the downregulation of miR-204 in glioma cells.Interestingly, miR-204 was an intronicmiRNA located betweenexons 7 and 8 of the TRPM3 gene (30). Consistent with aprevious report (30), expression of mature miR-204 and pri-miR-204 linearly correlated with that of TRPM3 in NHA andseveral glioma cell lines (Fig. 6A), indicating that miR-204 andTRPM3 might share common regulatory mechanism(s). Toexamine the epigenetic modification of this shared promoterregion, we retrieved the CpG-rich loci ofTRPM3 promoter fromthe UCSC genome browser (Fig. 6B) and assessed for their

methylation status via bisulfite genomic-sequencing PCR(BSP). The BSP result showed that in NHA, as well as in gliomacells with relatively higher levels of miR-204 expression, theCpG islands were mostly not methylated (Fig. 6C). By contrast,in genomic DNA obtained from glioma cells with downregu-lated miR-204 and from clinical glioma tumor samples, thepromoter CpG islands of miR-204 were hypermethylated inmost glioma specimens, strongly suggesting an essential role ofpromoter methylation inmiR-204 downregulation (Fig. 6C andSupplementary Fig. S4). To further validate this finding, wetreated the cells with DNA methyltransferase inhibitor 5-aza-dC for 72 hours and found that, in cells with low levels of miR-204 expression, inhibition of DNA methylation significantlypromoted the expression of miR-204 whereas, in NHA andglioma cells withoutmiR-204 promoter hypermetylation, miR-204 expression was not affected (Fig. 6D). Correspondingly, 5-aza-dC treatment could induce downregulation of SOX4 andEphB2 expression in glioma cell lines examined with low basal

Figure 5. miR-204 targets andsuppresses SOX4 and EphB2expression. A, conserved miR-204binding sites in SOX4 and EphB2generated by Targetscan database.B, Western blotting analysis ofexpression of SOX4 and EphB2 inindicated cells. a-Tubulin was usedas a loading control. C, relativeactivity of reporter luciferase linked toSOX4 or EphB2 30UTR measuredfollowing miR-204 mimic or miR-204inhibitor transfection in indicatedcells. D, enrichment of SOX4 andEphB2 30UTR in miRNP detected byqPCR following immunoprecipitationagainst Ago1. E, representativemicrographs of neural spheresformed by indicated cells after 7 daysof culture (left). Statistical analysis ofthe average diameter of spheres(right). F, representative pictures ofpenetrated cells analyzed using theTranswell matrix penetration assay.Error bars represent mean � SDobtained from 3 independentexperiments; �, P < 0.05;��, P < 0.01.






levels of miR-204 expression but could not suppress theexpression of SOX4 and EphB2 in cells with high levels ofendogenous miR-204 (Supplementary Fig. S5). Taken together,our data suggested that, in glioma cells, miR-204 downregula-tion was attributable to promoter methylation.

DiscussionDevelopment of stem cell-like properties in glioma cells has

been recognized as a hallmark contributing to disease pro-gression and generation of cellular population(s) robustlyempowered to migrate and invade the surrounding braintissue, to be resistant to chemo- or radiotherapies, resultingin frequent tumor recurrence, and to proliferate and colonizefrom a small number of initiating cells, in addition to otheraspects of tumor malignancies (31). Accumulated evidencesupports the notion that such a process is regulated bysignaling pathways similar to those governing the stemnessof neural stem cells. Notably, activation of these pathwaysrequires coordinated expression and cooperative action ofseveral core regulators. For example, SOX4 has recently beenidentified as a determinant factor of glioma stemness, whichforms a complex with OCT4 and subsequently activates thetranscription of SOX2, 1 of the core pluripotency reprogram-ming factors (24, 32). This transcriptional axis is essential forthe maintenance of glioma stem cell population and, moreimportantly, blocking this pathway greatly improves animalsurvival in a glioma xenograftmodel (24). As a short half-life (<1hour) protein (33), the expression level of SOX4 is tightly

controlled transcriptionally and posttranscriptionally. Inendometrial cancer, downregulation of miR-129-2 mediatesthe overexpression of SOX4, suggesting a significance ofmiRNA-mediated regulation in sustaining the expression ofthis stemness-associated oncoprotein (34). In glioma cells,however, the level of miR-129 has not been reported orscreened out in high-throughput assays to have significantchange. In this study, we identified miR-204 as a miRNAcommonly downregulated in both types of neural cells. Wethen further determined the targets of miR-204 and the con-sequent modulation of stem cell-like properties of glioma cells.Our data showed that miR-204 acts to inhibit the expression ofSOX4 and subsequently suppress the stem cell-like phenotypeof glioma cells, and the miR–204-downregulated status inglioma cells leads to generation of a glioma cell populationin which marked stemness is developed. In miR–204-trans-duced cells, ectopic expression of SOX4 is able to rescue neuralsphere-formation ability attenuated by miR-204, confirmingthat SOX4 is a major target of miR-204 and responsible formaintaining the stem cell population in human gliomas.Notably, miR-204 not only suppresses self-renewal but alsomigration of glioma cells, which is another hallmark of neuralstem cells (5). This study has also identified EphB2 as a majormigration regulator directly targeted by miR-204. Previously,EphB2 has been reported to promote stem cell migration inboth physiological and injury-repairing conditions (5, 26).Elevated expression and enhanced phosphorylation of EphB2has also been found in glioma cells (29, 35). More importantly,EphB2 has identified to be a critical promoter of tumor

Figure 6. Promoter methylation isresponsible for the loss of miR-204expression in glioma cells. A,relative expression ofmiR-204 (left)or pri-miR-204 (right) was plottedagainst expression of TRPM3accessed by qPCR. B, bisulfitegenomic sequencing resultingfrom 3 individual clones revealsmethylation status of promoterCpG of TRPM3 in indicatedspecimens. C, schematicillustration of the spatialarrangement of the promoterregion and miR-204 stem loop inTRPM3 genomic sequence. D,qPCR analysis of miR-204expression in indicated cellstreated with 5 mmol/L 5-aza-dC for72 hours. Error bars representmean � SD obtained from 3independent experiments; �, P <0.05; ��, P < 0.01.

Ying et al.





migration in glioma through interacting with the FAK andactivating FAK-mediated signal (28). The mechanism regulat-ing EphB2 expression in glioma cells, however, is still unclear.Our data thus shows a novel posttranscriptional regulator ofEphB2, namely, miR-204, epigenetic silencing of which enableselevation of EphB2 in glioma cells and enhances EphB2-medi-ated migration and invasion.Sustained activation of regulatory programs shared by gli-

oma and neural stem cells requires orchestrated transcriptionand posttranscriptional regulation of gene expression. Becauseof themultitarget property and the network effect, miRNAs areof great importance in the suppression of pluripotency ofembryonic stem cells by targeting reprogramming factors andinitiating differentiation (9). During the process of differenti-ation, an optimal miRNA expression pattern enables properdevelopment of the organism (36). In tumor cells, however,many of the stem cell-specific miRNA expression patterns arerearchived, such as the loss of neural-specific miR-124 expres-sion that promotes the stemness of glioma (15, 37). Consis-tently, wefind thatmiR-204, which is downregulated, is therebyexpected to play significant roles in both glioma cells andneural stem cells. miR-204 downregulation not only promotescellular stemness but also the biological properties essential tothe lethality of the disease, namely, migration and invasionabilities of glioma cells, both in vitro and in vivo. Most impor-tantly, restoration of miR-204 in glioma cells greatly abrogatesthe aggressiveness of glioma cells.Lastly, we showed that reduction of miR-204 is attributed to

downregulated transcription due to methylation of the pro-moter of it host gene TRPM3. TRPM3 is a member of thetransient receptor potentialmelastatin (TRPM) family that hasbeen reported to be associatedwith cancer progression. Down-regulation of TRPM1/melastatin mRNA in the primary cuta-neous tumor is a prognostic marker for metastasis in patientswith localized malignant melanoma (38). Moreover, TRPM8 islost in prostate cancer tissues from patients treated preoper-atively with antiandrogen therapy, suggesting that loss ofTRPM8 is associated with amore advanced form of the disease(39). Despite these findings, however, the pathobiological roleof TRPM3 in cancer remains largely unknown. In this context,

our currentfinding that the promoter ofTRPM3 is hypermethy-lated and TRPM3 mRNA is significantly downregulated ingliomas warrants further studies on the potential tumor-suppressive function of the TRPM3 gene in glioma. Moreover,it also remains highly interesting whether TRPM3 andmiR-204might cooperate with each other in the pathogenesis of glioma.Taken together, our present study provides new insights inunderstanding the stemness phenotype developed by gliomacells and, therefore, might contribute to future development ofnovel antiglioma strategies.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: Z. Ying, M-F LiDevelopment of methodology: Z. Ying, Y. Li, X. Zhu, Y. Yang, H. TianAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Z. Ying, Y. Li, J. Wu, H. Tian, W. LiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Z. Ying, Y. Li, J. Wu, X. Zhu, Y. Yang, H. Tian, W. LiWriting, review, and/or revision of themanuscript: Z. Ying, H. Tian, M-F Li,B. Hu, S-Y. ChengAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructingdatabases):Y. Li, J.Wu, X. Zhu, Y. Yang,H. Tian,M-FLiStudy supervision: M-F LiOther (produced and provided key reagents for this study): B. Hu, S-Y.Cheng

Grant SupportThis work was supported by the Ministry of Science and Technology of China

grant (grant no. 973-2011CB11305), the Natural Science Foundation of China(Grant nos. 81071647, 81071762, 30900415, 81272339, and 81272417), the Guang-dong Recruitment Program of Creative Research Groups (grant nos. 2009010058and 2010B030600003), the National Science and Technique Major Project (grantnos. 201005022-2, 2012ZX10004213, and 311030), the Key Science and TechniqueResearchProject ofGuangzhouProvince (grant no. 12B292060029); United StatesNational Institutes of Health (grant nos. CA130966 and CA158911), a scholaraward from the Zeller Family Foundation, and a research award for Brain Cancerfrom the James S. McDonnell Foundation.

The costs of publication of this article were defrayed in part by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 24, 2012; revised November 15, 2012; accepted November 17,2012; published OnlineFirst November 29, 2012.

References1. WenPY,Kesari S.Malignant gliomas in adults. NEngl JMed2008;359:

492–507.2. Sanai N, Alvarez-Buylla A, Berger MS. Neural stem cells and the origin

of gliomas. N Engl J Med 2005;353:811–22.3. Ben-Porath I, Thomson MW, Carey VJ, Ge R, Bell GW, Regev A, et al.

An embryonic stem cell-like gene expression signature in poorlydifferentiated aggressive human tumors. Nat Genet 2008;40:499–507.

4. Pedersen PH,Marienhagen K,Mork S, Bjerkvig R.Migratory pattern offetal rat brain cells and humanglioma cells in the adult rat brain. CancerRes 1993;53:5158–65.

5. CayreM,Canoll P,GoldmanJE. Cell migration in thenormal and patho-logical postnatal mammalian brain. Prog Neurobiol 2009;88:41–63.

6. Galli R, Binda E, Orfanelli U, Cipelletti B, Gritti A, De Vitis S, et al.Isolation and characterization of tumorigenic, stem-like neural pre-cursors from human glioblastoma. Cancer Res 2004;64:7011–21.

7. Dirks PB. Cancer: stem cells and brain tumours. Nature 2006;444:687–8.

8. Hammond SM. MicroRNAs as oncogenes. Curr Opin Genet Dev2006;16:4–9.

9. Xu N, Papagiannakopoulos T, Pan G, Thomson JA, Kosik KS. Micro-RNA-145 regulates OCT4, SOX2, and KLF4 and represses pluripo-tency in human embryonic stem cells. Cell 2009;137:647–58.

10. Jiang L, Lin C, Song L, Wu J, Chen B, Ying Z, et al. MicroRNA-30e�

promotes human glioma cell invasiveness in an orthotopic xenotrans-plantationmodel by disrupting theNF-kappaB/IkappaBalpha negativefeedback loop. J Clin Invest 2012;122:33–47.

11. Liu L, Wu J, Ying Z, Chen B, Han A, Liang Y, et al. Astrocyte elevatedgene-1 upregulates matrix metalloproteinase-9 and induces humanglioma invasion. Cancer Res 2010;70:3750–9.

12. https://tcga-data.nci.nih.gov/tcga/dataAccessMatrix.htm.13. http://www.ncbi.nlm.nih.gov/geo/.14. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, et al. TM4: a

free, open-source system for microarray data management and anal-ysis. Biotechniques 2003;34:374–8.






15. Silber J, Lim DA, Petritsch C, Persson AI, Maunakea AK, Yu M, et al.miR-124 and miR-137 inhibit proliferation of glioblastoma multiformecells and induce differentiation of brain tumor stem cells. BMC Med2008;6:14.

16. Srinivasan S, Patric IR, Somasundaram K. A ten-microRNA expres-sion signature predicts survival in glioblastoma. PLoS ONE 2011;6:e17438.

17. Chan JA, Krichevsky AM, Kosik KS. MicroRNA-21 is an antiapoptoticfactor in human glioblastoma cells. Cancer Res 2005;65:6029–33.

18. Corsten MF, Miranda R, Kasmieh R, Krichevsky AM, Weissleder R,Shah K. MicroRNA-21 knockdown disrupts glioma growth in vivo anddisplays synergistic cytotoxicity with neural precursor cell delivered S-TRAIL in human gliomas. Cancer Res 2007;67:8994–9000.

19. Papagiannakopoulos T, Shapiro A, Kosik KS. MicroRNA-21 targets anetwork of key tumor-suppressive pathways in glioblastoma cells.Cancer Res 2008;68:8164–72.

20. Kim TM, Huang W, Park R, Park PJ, Johnson MD. A developmentaltaxonomy of glioblastoma defined and maintained by microRNAs.Cancer Res 2011;71:3387–99.

21. Schraivogel D, Weinmann L, Beier D, Tabatabai G, Eichner A, Zhu JY,et al. CAMTA1 is a novel tumour suppressor regulated by miR-9/9� inglioblastoma stem cells. EMBO J 2011;30:4309–22.

22. Gabriely G, Yi M, Narayan RS, Niers JM, Wurdinger T, Imitola J, et al.Human glioma growth is controlled by microRNA-10b. Cancer Res2011;71:3563–72.

23. Sasayama T, Nishihara M, Kondoh T, Hosoda K, Kohmura E. Micro-RNA-10b is overexpressed in malignant glioma and associatedwith tumor invasive factors, uPAR and RhoC. Int J Cancer 2009;125:1407–13.

24. Ikushima H, Todo T, Ino Y, Takahashi M, Miyazawa K, Miyazono K.Autocrine TGF-beta signaling maintains tumorigenicity of glioma-ini-tiating cells through Sry-related HMG-box factors. Cell Stem Cell2009;5:504–14.

25. Bergsland M, Werme M, Malewicz M, Perlmann T, Muhr J. Theestablishment of neuronal properties is controlled by Sox4 and Sox11.Genes Dev 2006;20:3475–86.

26. Conover JC, Doetsch F, Garcia-Verdugo JM, Gale NW, YancopoulosGD, Alvarez-Buylla A. Disruption of Eph/ephrin signaling affectsmigra-tion and proliferation in the adult subventricular zone. Nat Neurosci2000;3:1091–7.

27. Holmberg J, Genander M, Halford MM, Anneren C, Sondell M, Chum-ley MJ, et al. EphB receptors coordinate migration and proliferation inthe intestinal stem cell niche. Cell 2006;125:1151–63.

28. Wang SD, Rath P, Lal B, Richard JP, Li Y, Goodwin CR, et al. EphB2receptor controls proliferation/migration dichotomy of glioblastomabyinteracting with focal adhesion kinase. Oncogene 2012 Feb 6. [Epubahead of print].

29. NakadaM, Niska JA,Miyamori H,McDonoughWS,Wu J, Sato H, et al.The phosphorylation of EphB2 receptor regulates migration and inva-sion of human glioma cells. Cancer Res 2004;64:3179–85.

30. Courboulin A, PaulinR,GiguereNJ,SaksoukN,Perreault T,Meloche J,et al. Role formiR-204 in humanpulmonary arterial hypertension. J ExpMed 2011;208:535–48.

31. Chen J, McKay RM, Parada LF. Malignant glioma: lessons fromgenomics, mouse models, and stem cells. Cell 2012;149:36–47.

32. Ikushima H, Todo T, Ino Y, Takahashi M, Saito N, Miyazawa K, et al.Glioma-initiating cells retain their tumorigenicity through integration ofthe Sox axis and Oct4 protein. J Biol Chem 2011;286:41434–41.

33. Beekman JM, Vervoort SJ, Dekkers F, van Vessem ME, VendelboschS, Brugulat-Panes A, et al. Syntenin-mediated regulation of Sox4proteasomal degradationmodulates transcriptional output.Oncogene2011;31:2668–79.

34. Huang YW, Liu JC, Deatherage DE, Luo J, Mutch DG, Goodfellow PJ,et al. Epigenetic repression of microRNA-129-2 leads to overexpres-sion of SOX4 oncogene in endometrial cancer. Cancer Res 2009;69:9038–46.

35. NakadaM, Niska JA, Tran NL,McDonoughWS, BerensME. EphB2/R-Ras signaling regulates gliomacell adhesion, growth, and invasion. AmJ Pathol 2005;167:565–76.

36. Pasquinelli AE, Ruvkun G. Control of developmental timing by micro-rnas and their targets. Annu Rev Cell Dev Biol 2002;18:495–513.

37. Gao FB. Context-dependent functions of specific microRNAs in neu-ronal development. Neural Dev 2010;5:25.

38. Duncan LM, Deeds J, Cronin FE, Donovan M, Sober AJ, Kauffman M,et al. Melastatin expression and prognosis in cutaneous malignantmelanoma. J Clin Oncol 2001;19:568–76.

39. Henshall SM, Afar DE, Hiller J, Horvath LG, Quinn DI, Rasiah KK, et al.Survival analysis of genome-wide gene expression profiles of prostatecancers identifies new prognostic targets of disease relapse. CancerRes 2003;63:4196–203.

Ying et al.





Published OnlineFirst November 29, 2012.Cancer Res Zhe Ying, Yun Li, Jueheng Wu, et al. Stem Cell-like PhenotypeLoss of miR-204 Expression Enhances Glioma Migration and

Updated version

10.1158/0008-5472.CAN-12-2895doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2012/11/29/0008-5472.CAN-12-2895.DC1

Access the most recent supplemental material at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/early/2013/01/10/0008-5472.CAN-12-2895To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-12-2895

http://cancerres.aacrjournals.org/content/suppl/2012/11/29/0008-5472.CAN-12-2895.DC1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2013/01/10/0008-5472.CAN-12-2895


loss of mir-204 expression enhances glioma - cancer research

Documents