TheOncogenicRNA-BindingProteinMusashi1IsRegulatedby HuR via mRNA Translation … · Musashi1(Msi1)isanevolutionarilyconservedRNA-bindingprotein(RBP)thathasprofoundimplicationsin cellular

Signaling and Regulation

The Oncogenic RNA-Binding Protein Musashi1 Is Regulated byHuR via mRNA Translation and Stability in Glioblastoma Cells

Dat T. Vo1,2, Kotb Abdelmohsen6, Jennifer L. Martindale6, Mei Qiao1, Kumiko Tominaga6, Tarea L. Burton1,Jonathan A.L. Gelfond3, Andrew J. Brenner4,5, Vyomesh Patel7, Daniel Trageser8, Bj€orn Scheffler8,Myriam Gorospe6, and Luiz O.F. Penalva1,2

AbstractMusashi1 (Msi1) is an evolutionarily conserved RNA-binding protein (RBP) that has profound implications in

cellular processes such as stem cell maintenance, nervous system development, and tumorigenesis. Msi1 is highlyexpressed in many cancers, including glioblastoma, whereas in normal tissues, its expression is restricted to stemcells. Unfortunately, the factors that modulate Msi1 expression and trigger high levels in tumors are largelyunknown. TheMsi1mRNAhas a long 30 untranslated region (UTR) containing several AU- andU-rich sequences.This type of sequence motif is often targeted by HuR, another important RBP known to be highly expressed intumor tissue such as glioblastoma and to regulate a variety of cancer-related genes. In this report, we show aninteraction betweenHuR and theMsi1 30-UTR, resulting in a positive regulation ofMsi1 expression.We show thatHuR increased MSI1 mRNA stability and promoted its translation. We also present evidence that expression ofHuR and Msi1 correlate positively in clinical glioblastoma samples. Finally, we show that inhibition of cellproliferation, increased apoptosis, and changes in cell-cycle profile as a result of silencing HuR are partially rescuedwhenMsi1 is ectopically expressed. In summary, our results suggest that HuR is an important regulator of Msi1 inglioblastoma and that this regulation has important biological consequences during gliomagenesis.Mol Cancer Res;10(1); 143–55. �2012 AACR.

Introduction

Cancer is a disease process ultimately triggered by aberrantgene expression. In this scenario, there is a large body ofliterature discussing the participation of transcription andchromatin modification factors in tumorigenesis. On theother hand, the role of posttranscriptional regulation intumor growth is still poorly comprehended. RNA-bindingproteins (RBP) are key mediators of posttranscriptional generegulation, mediating many aspects of RNA metabolismsuch as splicing, transport, localization, translational regu-lation, and stability (1). In addition to controlling normalprocesses in cellular physiology, aberrant expression of RBPs

has been shown to lead to tumorigenesis when their targetgenes are implicated in cell proliferation, differentiation,apoptosis, invasion, and metastasis.An RBP of particular interest in cancer is Musashi1

(Msi1), first described as a gene required for the developmentof the Drosophila adult external sensory organ (2). Inmammals, Musashi1 is required for nervous system devel-opment during embryonic development, whereas in adults,Musashi1 expression is mainly restricted to stem andprogenitor cells of various tissues (3–6). Aberrantly highMusashi1 expression is observed in many cancers such asmedulloblastoma (7), hepatocellular carcinoma (8), cervicaladenocarcinoma (9), lung cancer (10), colon cancer (11),and glioblastoma multiforme (GBM). In fact, increasingexpression of Musashi1 has been correlated with a poorprognosis in glioma (12), breast cancer (13), and medullo-blastoma (Penalva Lab, unpublished data).During normal development, Musashi1 maintains stem

cell identity, serving as a key gene in stemness (14).However,in the tumor environment, Musashi1 enhances cancerfeatures. High Msi1 expression is associated with increasedNotch1 expression and areas of tumor invasion/metastasis(12). Notch is a critical pathway for tumorigenesis inmedulloblastoma, glioblastoma, and other tumor types;Msi1 influences this pathway by repressing the regulationofNUMBmRNA, a negative regulator ofNotch (15–17). Inmedulloblastoma, Msi1 inhibition results in increased sen-sitivity to the Hedgehog pathway inhibitor cyclopamine,

Authors' Affiliations: 1Greehey Children's Cancer Research Institute,Departments of 2Cellular and Structural Biology, 3Epidemiology and Bio-statistics, 4Division of Hematology and Medical Oncology, Department ofMedicine, and 5Cancer Therapy & Research Center, University of TexasHealth Science Center at San Antonio, San Antonio, Texas; 6Laboratory ofMolecular Biology and Immunology,National Institute onAging - IntramuralResearch Program, NIH, Baltimore; 7Oral and Pharyngeal Cancer Branch,National Institute of Craniofacial and Dental Research, NIH, Bethesda,Maryland; and 8Institute of Reconstructive Neurobiology, University ofBonn Medical Center, Bonn, Germany

Corresponding Author: Luiz O.F. Penalva, Department of Cellular andStructural Biology, Greehey Children's Cancer Research Institute, Univer-sity of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive,SanAntonio, TX 78229. Phone: 210-562-9049; Fax: 210-562-9014; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-11-0208

�2012 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 143

on August 5, 2021. © 2012 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

http://mcr.aacrjournals.org/

indicating that Msi1 interfaces with the Hedgehog pathway(18). In murine in vivo xenograft experiments using breastand colon cancer cells, silencing ofMsi1 via siRNAs results ininhibition of tumor growth (13, 19); similar results wereobserved for glioblastoma and medulloblastoma cells(Penalva Lab, unpublished data). Our results indicate thatMsi1 influences tumor progression in a complex manner byregulating the expression of a network of genes implicated incancer-related processes such as cell proliferation, apoptosis,cell cycle, and differentiation (17).As summarized above, a large amount of data support the

role of Msi1 as an oncogenic protein. However, the molec-ular determinants of increased Musashi1 expression duringtumorigenesis are largely unknown. In normal stem cells,one study identified the HuD RBP as a potential posttran-scriptional regulator of Musashi1 expression, aiding neuralstem cells in the transition toward differentiation (20). It hasalso been suggested that a potential regulatory element at thetranscriptional level exist as evident by the presence of ahypoxia-responsive element which can bind the hypoxia-inducible factor 1 in times of hypoxic stress, promoting self-renewal and proliferation in neural stem cells (21). However,neither of these elements can explicate the overexpression ofMusashi1 during tumorigenesis. The mRNA of Msi1 con-tains a long 30 untranslated region (UTR), spanning approx-imately 1,800 nucleotides, making it a likely candidate forposttranscriptional regulation. We have recently shown thatMsi1 expression is regulated by several tumor suppressormicroRNAs (22). Furthermore, the 30-UTR contains severalsegments of AU- or U-rich cis-regulatory sequences, makingit a potential target of turnover and translation regulatoryRBP (TTR-RBP) such as HuR (also known as ELAVL1 orHuA), a member of the Hu/ELAV (embryonic lethal abnor-mal vision) family (23).HuR can enhance tumorigenesis by interacting with a

subset of mRNAs that encode proteins that regulate cellproliferation, cell survival, angiogenesis, invasion, andmetas-tasis (24). Many studies have reported elevated expression ofHuR in numerous malignancies (25, 26). HuR contains 3RNA recognition motifs that bind to 30-UTRs of mRNAsthat bear AU- orU-rich sequences (27). Inmalignant tumorsof the central nervous system such as glioblastoma, HuR hasbeen linked to the augmented expression of genes such asTNF-a, interleukin (IL)-8, COX2, VEGF, TGF-b, andother genes involved in enhancing the tumorigenic pheno-type, such as increased cell proliferation, evasion of apopto-sis, angiogenesis, and invasion/metastasis (25, 26, 28). Amore recent study showed that HuR promotes an antia-poptotic phenotype through the posttranscriptional controlof Bcl-2, Mcl-1, and Bcl-xl, proteins belonging to the Bcl-2family of antiapoptotic proteins (29).In this study, we show that high expression of Msi1 in

glioblastoma is partially triggered by HuR through itsinfluence onMsi1 translation andmRNA stability, resultingin an increased steady-state level of MSI1 mRNA andincreased Msi1 protein output. Supporting this idea, weobserved that HuR and Msi1 have similar patterns ofexpression. Finally, we showed that Msi1 transgenic expres-

sion overcomes the impact of HuR knockdown on cellproliferation, apoptosis, and cell-cycle profile. In conclusion,we suggest that the HuR-Msi1 link is an important piece ingliomagenesis.

Materials and Methods

Cell cultureU251 and U343 glioblastoma cells were maintained in

Dulbecco'sModified EssentialMedium (Thermo Scientific)supplemented with 10% FBS, penicillin, and streptomycin.HeLa cervical adenocarcinoma cells were maintained inMinimum Essential Medium (Thermo Scientific) supple-mented with 10% FBS, penicillin, and streptomycin. Pri-mary glioblastoma tumorspheres were obtained from surgi-cally resected patient tumors and propagated in neurobasalmedia containing L-glutamine, N2 supplement (Gibco),B27 supplement (Gibco), heparin (Sigma), epidermalgrowth factor (EGF; PeproTech, Inc.), and basic fibroblastgrowth factor (bFGF; PeproTech, Inc.; ref. 30). For growthof primary glioblastoma cells as monolayers, the cells werecultured in the presence of Dulbecco's Modified EssentialMedium with 10% FBS, penicillin, and streptomycin. TheU251 overexpression cell lines were previously established(22). U343 overexpression cell lines were created throughstable G418 (Gibco) selection (800 mg/mL) after transfec-tion of the pEF1/myc-His A plasmid (Invitrogen), whichcontains the expression cassette encoding for an EF-1apromoter–driven coding region for eitherMsi1 or GFP. Afterinitial selection, 400 mg/mL of G418 was used for mainte-nance. The levels of Msi1 overexpression were monitoredusing reverse transcription quantitative PCR (qRT-PCR)analysis.

Immunoprecipitation of HuR ribonucleoproteincomplexesImmunoprecipitation of HuR ribonucleoprotein (RNP)

complexes (31) was carried out using U251 glioblastoma celllysates. Cell lysates were prepared and precleared using 15 mgof IgG isotype control and protein A-sepharose beads, pre-swollen in NT2 buffer (50 mmol/L Tris-HCl, pH 7.4, 150mmol/L NaCl, 1 mmol/L MgCl2, 5% bovine serum albu-min, and 0.05%Nonidet P-40), for 30minutes at 4�C. Afterpreclearing, 100 mL of protein A-sepharose beads wereincubated with 30 mg of anti-HuR antibody for 18 hoursat 4�C then for 1 hour at room temperature with 3 mg ofprecleared cell lysate. After repeated washing and centrifu-gation of the sepharose beads with ice-cold NT2 buffer,HuR-boundRNAwas released fromprotein using proteinaseK and SDS. RNA was extracted using acid phenol–chloro-form and ethanol precipitated. After reverse transcription(RT), qPCR analysis was carried out using gene-specificprimers for MSI1, GAPDH, and PTMA mRNAs.

In vitro transcription of biotinylated RNAand analysis ofHuR bound to biotinylated RNAA plasmid containing full-length Msi1 cDNA was used as

a template for PCR. The PCR products were purified and

Vo et al.

Mol Cancer Res; 10(1) January 2012 Molecular Cancer Research144



used as a template for in vitro transcription. In vitro tran-scription was catalyzed by T7 RNA polymerase, in thepresence of biotin-CTP. Biotin pull-down assays were car-ried out by incubating 40 mg of cytoplasmic fractions with 1mg of biotinylated transcripts for 1 hour at room tempera-ture. Complexes were isolated with paramagnetic streptavi-din-conjugated Dynabeads (Dynal), and bound proteinswere analyzed by Western blotting by using a mouse mono-clonal antibody recognizing HuR. After secondary antibodyincubations, signals were visualized by chemiluminescence.

Reporter assayFor luciferase assays, HeLa cells were cotransfected with

the pSGG-Msi1 reporter vector and TAP-tagged HuRexpression vector with Roche Fugene 6 transfection reagent(Roche Applied Sciences). An empty expression vector wasused as a negative control. A total of 8� 103HeLa cells wereplated in a 96-well culture plate 24 hours prior to transfec-tion. Using the Fugene 6 reagent, 100 ng of the HuRexpression vector (or empty vector for a negative control)and 1 ng of the UTR luciferase reporter vector werecotransfected. Forty-eight hours later, cell lysates were pre-pared with the Reporter Lysis Buffer (Promega). Using theLuciferase Assay Reagent (Promega), luminescence was readon aBertholdTechnologies AutoLumat LB953Multi-Tubeluminometer (Berthold Technologies).

siRNA transfectionsU251 and U343 glioblastoma cells were reverse trans-

fected with siRNAs with Lipofectamine RNAiMAX trans-fection reagent (Invitrogen). siRNAs targeting HuR wereobtained from the Dharmacon RNAi Collection SMART-pool Line (Thermo Scientific). Negative control siRNAreagent was obtained from Thermo Scientific Dharmacon.U251 and U343 glioblastoma cells were transfected withsiRNAs in 100-mm culture plates for Musashi1 and HuRprotein analysis. Seventy-two hours posttransfection, cellswere scraped from the plate and harvested for analysis.

Quantitative RT-PCRTotal RNA was extracted using the TRIzol reagent (Invi-

trogen). After lysis with TRIzol and separation of any RNA:protein complexes, total RNA was extracted into the aque-ous phase using chloroform. Total RNAwas precipitated outof the aqueous phase using isopropanol, washed with 75%ethanol, and resuspended in nuclease-free water (Ambion).The cDNA was synthesized using Applied Biosystems

High Capacity cDNA Reverse Transcription Kit (AppliedBiosystems) using random priming. Quantitative PCR wascarried out using the TaqMan Gene Expression Master Mix(Applied Biosystems). All reactions were run on anABI 7500Real Time PCR machine (Applied Biosystems). ABI Taq-Man Gene Expression Assay (Applied Biosystems) primer/probe sets were obtained for the ELAVL1,MSI1, and ACTBmRNAs. The data were acquired using the ABI SDS 2.0.1software package. ACTBmRNAwas used as an endogenouscontrol, and the data were analyzed using the 2 �DDCtð Þmethod.

Western blotAfter collection of cells for analysis, cells were resuspended

and sonicated in 2� SDS Laemmli sample buffer. A 10%SDS-PAGE gel with a 4% stacking gel was run in Tris-glycine-SDS buffer. A semi-dry transfer procedure wascarried out onto a nitrocellulose membrane. After transfer,the membrane was blocked in TBS with Tween-20 and 5%milk. The membrane was probed with a rabbit monoclonalanti-Musashi1 antibody (Abcam), a mouse monoclonalanti-HuR antibody (Abcam), or mouse monoclonal anti-a-tubulin antibody (Sigma). Horseradish peroxidase(HRP)-conjugated goat anti-rabbit antibody (Santa CruzBiotechnology) was used as a secondary antibody for Musa-shi1 or HRP-conjugated goat anti-mouse antibody (ZymedLaboratories) was used as a secondary antibody for HuR anda-tubulin. Electrochemiluminescence was used to detect theMusashi1, HuR, and a-tubulin protein using the Super-Signal West Pico Chemiluminescent Substrate (ThermoScientific, Pierce Protein Products).

mRNA stability assayU251 glioblastoma cells were reverse transfected with

either the anti-HuR siRNA or scrambled negative controlsiRNA, in 35-mm plates at 50% confluency in antibiotic-free media. After 12 hours, the media were switched tomedia containing 5 mg/mL actinomycin D (Alexis Bio-chemicals) to inhibit transcription; each hour after that,total RNA was harvested using TRIzol (Invitrogen) fol-lowing the manufacturer's protocol. The cDNA wasreverse transcribed using the ABI High Capacity cDNAReverse Transcription Kit (Applied Biosystems). TaqManreal-time PCR was carried out on the cDNA using primers(Applied Biosystems) for MSI1 and GAPDH mRNA.Quantitative PCR was carried out on an ABI 7500 RealTime PCR system (Applied Biosystems), and data wereacquired using the ABI SDS 2.0.1 Software Package(Applied Biosystems) and analyzed using the 2 �DDCtð Þmethod. Subsequently, the data were fitted to a linearregression, and the regression was used to calculate thehalf-life of the MSI1 mRNA.

Polysomal gradient preparation and analysisHuR siRNAs, or corresponding negative control siRNA,

were reverse transfected into U251 glioblastoma cells. Sev-enty-two hours after transfection, translation was arrestedusing 0.1 mg/mL of cycloheximide. Cells were dissociatedusing trypsin, and cell pellets were formed. Pellets wereresuspended and lysed in polysome extraction buffer (20mmol/L Tris-HCl, pH 7.5, 100 mmol/L KCl, 5 mmol/LMgCl2, 0.3% Igepal CA-630, protease inhibitors, and 0.1mg/mL cycloheximide). After centrifugation to removeinsoluble material, the lysate was overlaid on a 10% to50% sucrose gradient in a buffer solution of 20 mmol/LTris-HCl, pH 7.5, 100 mmol/L NaCl, and 5 mmol/LMgCl2. After ultracentrifugation at 39,000 rpm at 4�C,1-mL fractions were obtained on a density gradient frac-tionation system (Brandel). A254 was measured during theentire fractionation process. Total RNA was extracted from

HuR Regulation of Musashi1

www.aacrjournals.org Mol Cancer Res; 10(1) January 2012 145



each fraction using TRIzol (Invitrogen) and analyzed byreverse transcription followed by qPCR analysis.

Quantitative RT-PCR of clinical glioblastoma samplesPreviously characterized GBM specimens were used for

this study (32). Briefly, tissue samples were obtained from thesolid tumor, minced, and enzymatically dissociated. Primarytumor cell cultures weremaintained under serum-free condi-tions according to Lee and colleagues (33). Clinical gliobla-stoma sampleswere prepared for total RNAusing theRNeasyMini Kit (Qiagen). cDNA was prepared using the ABI HighCapacity cDNA Reverse Transcription Kit (Applied Bio-systems), and real-time PCR reactions were carried out withTaqManGene Expression assay probes (Applied Biosystems)for MSI1, ELAVL1, and ACTB mRNA. The data wereacquired using the ABI SDS 2.0.1 Software Package (AppliedBiosystems) and analyzed using the 2 �DDCtð Þ method. TheqRT-PCR measurements of HuR/Msi1 were carried out intriplicate and averaged for each tumor sample (n ¼ 20) andthe control tissue sample U251. The Ct value of these geneswas normalized by subtracting the Ct of the endogenouscontrol ACTB mRNA measurement.

Tissue microarray and analysisImmunohistochemistry was conducted on the GL806

brain glioblastoma and normal tissue array, which contains35 cases of glioblastoma, 5 normal tissue, duplicated coresper case (US Biomax Inc.). Antigen retrieval was conductedby heat induction of the array slides. Monoclonal anti-HuRantibody (Molecular Probes Inc.) and polyclonal anti-Msi1antibody (R&D Systems, Inc.) were used for immunohis-tochemical staining of the tissue arrays.All stained slides were scanned at 40� using an Aperio CS

Scanscope (Aperio Technologies) and quantified using theavailable Aperio algorithms. Immunodetection of Msi1 orHuR was quantified according to percentage of cells stained.The percentages of stained cells were multiplied by 3 forintensively stained cells, moderately stained cells by 2, andmildly stained cell by 1 and used as a staining index for thequantitative assessment of changes in marker expression ineach treated group, as previously reported (34).

Cell proliferation assayCell proliferation was conducted on the U251 Msi1 or

GFP overexpression cell lines (U251:MSI1 or U251:GFP,respectively) and U343 Msi1 or GFP overexpression celllines (U343:MSI1 or U343:GFP, respectively). Either HuRsiRNA or control siRNA was transfected into each cell line.After 48 hours of transfection, cell proliferation was mea-sured with CellTiter 96 AQueous Non-Radioactive CellProliferation Assay (Promega). After 1 hour, absorbance at490 nm was measured on a Synergy HT Multi-Modemicroplate reader (Biotek).

Cell-cycle analysisCells were prepared for cell-cycle distribution analysis

using propidium iodide staining of DNA content. Cellswere harvested using trypsin, pelleted via centrifugation at

1,000 rpm for 3minutes at 4�C, and resuspended in ice-coldPBS. The cells were pelleted once again and resuspended in500 mL of ice-cold PBS. The resuspended cells were thenslowly dropped into a vortexing tube containing 9� volumeof ice-cold 100% ethanol for fixation. The solution wasincubated at 4�C for 30 minutes. The cells were thenpelleted via centrifugation, washed with PBS once more,and resuspended in the propidium iodide staining solution,which is composed of a solution of PBS containing 0.1%(v/v) of Triton-X-100 (Sigma-Aldrich), 200 mg/mL ofRNase A (Sigma-Aldrich), and 20 mg/mL of propidiumiodide (Invitrogen). The cells were incubated in stainingsolution in the dark for 30 minutes at room temperature.The cells were analyzed on the BD FACSCanto FlowCytometry Machine (BD Biosciences). Flow cytometricdata were analyzed using Modfit LT (Verity SoftwareHouse) to calculate cell-cycle distribution.

Caspase-3/7 activity assayApoptosis was assayed with the U251 Msi1 or GFP

overexpression cell lines (U251:MSI1 orU251:GFP, respec-tively) and U343 Msi1 or GFP overexpression cell lines(U343:MSI1 or U343:GFP, respectively). Either HuRsiRNA or control siRNA was transfected into each cell line.After 48 hours of transfection, apoptosis wasmonitored withthe Caspase-Glo 3/7 Assay Reagent (Promega). Luciferasemeasurements were acquired with the SpectraMax M5Multi-Mode Microplate Reader (Molecular Dynamics).

Statistical analysisAll statistical analysis was conducted with the Student t

test. Data are presented as the mean � SEM.

Results

Identification of U- and AU-rich elements in the Msi130-UTRTheMSI1 mRNA contains a long 30-UTR, suggestive of

posttranscriptional regulation. In a previous report, ourlaboratory has identified 5 tumor suppressor miRNAs thatmodulateMsi1 expression (22).Upon closer observation, wealso identified regions in the Msi1 30-UTR containingpotential uracil-rich elements amenable for regulation byHuR (ref. 35; Fig. 1A). In agreement with the idea that HuRfunctions as a positive regulator of Msi1, HuR has beenshown to be upregulated in many tumors (24) includingglioblastoma, where Msi1 is also highly expressed.

Immunoprecipitation of HuR enriches for the MSI1mRNAFirst, we conducted an immunoprecipitation of HuR

ribonucleoprotein complexes (RNP-IP) in U251 glioblas-toma cells followed by qRT-PCR to determine whether ornot the MSI1 mRNA is associated with HuR protein. In 3biological replicates, we observed an overall enrichment ofapproximately 20-fold of the MSI1 mRNA in immunopre-cipitation carried out with anti-HuR antibodies in compar-ison to the GAPDH mRNA, which does not bind to HuR.

Vo et al.




As a positive control for HuR binding, the mRNA encodingprothymosin-a, an antiapoptotic protein, was evaluated(36). HuR binding to MSI1 mRNA seems to be evenstronger than HuR binding to prothymosin-a mRNA(Fig. 1B).To map the region of theMSI1 mRNA that contains the

HuR-binding site, we carried out biotin pull-down experi-ments with biotinylated fragments of the MSI1 mRNA.Three fragments were made, one for the coding region ofMsi1 and two for the 30-UTR (a proximal region and a distalregion; Fig. 1C). The GAPDH 30-UTR was included as anegative control. After incubation with U251 cell lysate, thebound protein eluate was immunoblotted for HuR. Theresults of the pull-down Western blot experiments show astrongHuR signal for the distal half of theMsi1 30-UTR thanfor the beads only (Fig. 1D). Aweak but positive signal is seen

for the proximal region, indicating that indeedbinding exists,butmost likely with low affinity "and biotinylatedRNA" canbe removed as the fragments referred to in the beginning ofthe sentence refers to the biotinylated RNA. The codingregion fragment shows no signal, indicating a lack of abinding of HuR to this region of the MSI1 mRNA.

Luciferase reporter assay confirms interaction with Msi130-UTRTo further examine the effect of HuR on Msi1 expres-

sion, a luciferase reporter carrying the Msi1 30-UTRwas constructed and tested. The luciferase coding regioncontained a PEST degradation sequence (37) that effec-tively reduces the luciferase protein half-life to approxi-mately 1 hour, thus reducing the chances of any ambi-guous results due to protein accumulation. Compared

MSI1 3′-UTR

cap

poly (A)MSI1 3′ coding region

agcaggg gacggtggca ggagcgcccc agcctgcagc tgactgagga1201 ccacgagtga gccagcgagg gggcgggaga cctcagccgc agccgccgcc ccctcccctg1261 cagcgactcg gacccgctac tgcctgcccc caactccccg ggcccggccc ctgccctgct1321 gcccccaaca gcgtctggct cccctactaa cgtccccctc ttcgcccttg cccccatccc1381 cacccgcccc tctcccggcc ctgcttttat ttattttgga ttagccggtt gcccacccca1441 gccctctggt ccatccctcc ctccgtgccg cgccccccta ggaccgcccc ctccccaaaa1501 ggcttttgga tttgtgcata gctggagtga aggcggaggg agcctgctac aggccgcagc1561 ccaacccctg ttttttattc agatttcccc tcctttaccc tttccctttt tttttttttt1621 ttttttttta aagaaacctt ttttaaacta tttctaggtt tgtgaatgtg aagccccagg1681 ccgcaggggg caaggggcca ggtgcccccc accagctgag aacaaagtgt ctatctgggt1741 gtgggcccct ggccgcctcc ctccagccct ggagaggagg gcagggctgc ggggaggcca1801 ggccgagccc ctggaaccat cccgtcctgt atcatatgta aatactgtga ggtgatgtgc1861 ccacccctct ctaagacccc tcgggggtga ggggctcccc cctcccctgt ttctgtcccc1921 tcagacaccg ttactgtaag cttgcaggcc tcagctgtgg ccacggcagg cccgctctct1981 caggccctca gggtcaaggc cttggttgga cctgcccact ccaaaaaccc agtgtggggg2041 caaagggcgt gggaagagca gggcttgccc agcgacactg ctggacagga attaactctc2101 caaaggtctc ccctgctccc tacctaggtt ggggcttcat ggtttctgct cagtctgtcc2161 cccttccccc tcgacccccg caatgagtgg gcaccagggg acgctctggc gagggcagac2221 cccaggggaa agcaaagggc gtctcaggga acccccacat tctctcactg aagtttccca2281 ccaggatgac cccacagcca gagtcccttg gcagcccctc accccagacc ccccttctaa2341 ggaaaaagag aggttcagag cgttggacct tcatgaacag tggcctggct ggcgtggcag2401 ggccaaggcc cacccactgc catccccctt ctgtgtgtcc ctctccccca gttatgaggc2461 ccagggcctg agctctcttc cccagcattg cccccacccg gaaaccccac ctttggagag2521 ttaattgtct gtgtgaggtg cttaacctat cagccctgag aacacaaagc aataatcttt2581 gttactgaga tgcgcggctg ttcgtgtttt gggttttttt tttttttaat gtttcctaat2641 aaaagagaag ctgcatttta ttggttttta tttttaattt tctacacgtt tgagctgagt2701 cctgagacac ttagcttccc tctccctcat tcccggaccc ttccacccca ctgggcccaa2761 ccatgggctc aggaccctgg aattccgttt tctgatttgc tttgggattt tttttttttt2821 taagatgtta catggtgttt cgaagccagc aagttaccat cctccggtgt ctcttctctc2881 cacatctgta acttcttttt ccaggtttta ttttcagttt taaattccta ataaattatt2941 tgaaaacgtt

A

0

5

10

15

20

25

GAPDH PTMA MSI1 MSI1 MSI1

Rela

tive m

RN

A b

indin

g

B

D

Lysa

te

Be

ad

s

CD

S

3′-U

TR

-A

3′-U

TR

-B

GA

PD

H 3

′-UT

R

HuR

C

CDS 3′-UTR-A 3′-UTR-B

E

0

0.5

1

1.5

2

2.5

luc2:UTR luc2:UTRTAP TAP-HuR

P = 2.5 x 10-5

Re

lative

lucif

era

se

exp

ressio

n

Figure 1. Discovery of HuR as a trans-acting posttranscriptional regulatory factor for Msi1. A, a schematic of theMSI1mRNAwith the 30-UTR shown in the inset.The nucleotide tracts bolded, italicized, and underlined indicates the putative HuR-binding sites. B, immunoprecipitation of RNA:HuR protein complexesidentifies Musashi1 mRNA as a binding partner for HuR. HuR RNP complexes were pulled down with an anti-HuR–specific antibody. The RNA recovered wasreverse transcribed and qPCR was carried out on the cDNA. Primers were used for GAPDH, PTMA, and MSI1 mRNAs. GAPDH mRNA is used as a negativecontrol and PTMA mRNA is a positive control. C, diagram of the mRNA fragments used in biotin pull-down experiment. Three fragments were made, one forthe coding region (CDS) and two for each half of the 30-UTR. D, biotin pull-down assay confirms HuR interaction withMSI1 mRNA. Biotinylated fragments oftheMSI1mRNAwere in vitro transcribed. Three fragmentsof themRNAwere analyzed, one for the coding regionand two for eachhalf of the30-UTR.TheGAPDH30-UTR was included as a negative control. After incubation of the fragments with streptavidin beads, the protein was recovered by pull-down analysis andimmunoblotted for HuR. A beads only control was included to account for any nonspecific interaction and a "lysate" was included as a positive control for theimmunoblot. E, influence of HuR on Msi1 protein expression is dependent on the 30-UTR. The Msi1 30-UTR was cloned downstream of a PEST-destabilizedluciferase gene (luc2P). The luciferase construct was cotransfected with a TAP-tagged HuR expression vector (denoted as TAP-HuR) in HeLa cervicaladenocarcinoma cells. An empty vector (denoted as TAP) was used as a negative control. Data were analyzed with the Student t test and are presented as themean � SEM. Experiment was carried out in triplicate. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.





with the empty expression vector, we detect a 2-foldincrease (P ¼ 2.5 � 10�5) in luciferase activity whenwe cotransfect a HuR expression vector (Fig. 1E), thusconfirming the interaction of HuR with the Msi1 30-UTR.

HuR silencing suggests that HuR positively regulatesMsi1 expressionTo further validate the regulation of Msi1 by HuR, its

levels were reduced by using HuR-directed siRNA in U251andU343 glioblastoma cells. A significant reduction inHuR(U251: P ¼ 1.5 � 10�9; U343: P ¼ 1.3 � 10�3) wasfollowed by a concomitant reduction in Msi1 expression atboth the mRNA (P ¼ 3.4 � 10�7; U343: P ¼ 5.1 �10�3; Fig. 2A and B) and protein levels (Fig. 2C and D),suggesting that HuR positively regulates Msi1 expression inglioblastoma cells.

HuR increases MSI1 mRNA stabilityTo dissect the mechanism by which HuR controls Msi1

expression, we first measured the stability of MSI1 mRNA(38). We knocked down the expression of HuR using

siRNAs and 12 hours posttransfection, we treated the cellswith 5 mg/mL of actinomycin D, an inhibitor of RNApolymerase elongation. After transcription was inhibited,mRNA decay was measured without interference. ThemRNA decay rate of the MSI1 mRNA was measured inthe presence and absence of the HuR RBP. Over the courseof 3 hours, we observed a faster rate of decay inHuR-silencedcells than in control cells. We fitted the data to a linearregressionmodel and used themodel to calculate the half-lifeof the MSI1 mRNA. The half-life of the MSI1 mRNA is8.17 hours, whereas inHuR-silenced cells, theMSI1mRNAhalf-life was reduced to approximately 3.12 hours, repre-senting a 62% increase in the rate of decay (P¼ 4.8� 10�5)of the MSI1 mRNA (Fig. 3).

HuR positively regulates Msi1 translationIn addition to mediating mRNA stability, HuR can also

modulate translation (24). To determine whether HuR alsoaffects the translation of MSI1 mRNA, we conductedpolysome fractionation analysis using extracts from U251control and HuR-knockdown cells (39). In this assay,polysomes, which contain actively translating mRNAs andhave high sedimentation rates, are partitioned from untrans-lated mRNAs, which have low sedimentation rates (Fig. 4Aand B). Fine fractions are acquired from the gradients andlevels of mRNAs can be measured with Northern blot orqRT-PCR. The results showed that the profile of theGAPDH mRNA, encoding a housekeeping protein, was

HuR

Msi1

α-Tubulin

Control HuRsiRNA siRNA

A

C

P = 1.5 x 10-9

P = 3.4 x 10-7

0

0.2

0.4

0.6

0.8

1

1.2

Control siRNA

HuR siRNA

Re

lative

mR

NA

exp

ressio

n

HuR Msi1

0

0.2

0.4

0.6

0.8

1

1.2

Control siRNA

HuR siRNA

Re

lative

mR

NA

exp

ressio

n

HuR Msi1

P = 1.3 x 10-3

P = 5.1 x 10-3B

D

HuR

Msi1

α-Tubulin

Control HuRsiRNA siRNA

Figure 2. MSI1 mRNA and protein levels are dependent on HuR levels.A and B, MSI1 mRNA levels are upregulated by HuR. Seventy-twohours after siRNA-mediated silencing of HuR (50 nmol/L) in U251 (A)and U343 (B) glioblastoma cells, MSI1 mRNA was quantified byreverse transcription followed by qPCR analysis; b-actin mRNA wasused as an endogenous control and the data were analyzed usingthe 2 �DDCtð Þ method. Data were analyzed with the Student t test and arepresented as the mean � SEM. Experiment was carried out intriplicate. C and D, Msi1 protein is positively dependent on HuR levels.By 72 hours after siRNA-mediated silencing of HuR (50 nmol/L) inU251 (C) and U343 (D) glioblastoma cells, Msi1 protein levels wereassessed by Western blot analysis. a-Tubulin served as a loadingcontrol.

Control siRNA – t1/2 = 8.17 hHuR siRNA – t1/2 = 3.12 h

P = 4.8 x 10-5

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3

Rela

tive m

RN

A e

xpre

ssio

n

Time after actinomycin D addition (h)

Control siRNA HuR siRNA

Figure 3. HuR stabilizes theMSI1mRNA. The stability of theMSI1mRNAmediated by HuR was assessed in U251 glioblastoma cells. siRNAs (50nmol/L) directed against HuR were transfected into U251. A scrambledsiRNA was used as a negative control. Actinomycin D (5 mg/mL) wasadded to the cell media 12 hours after siRNA transfection. Total RNAsamples were obtained, and MSI1 mRNA levels were determined usingqRT-PCR. The data were analyzed using the 2 �DDCtð Þ method and theGAPDH mRNA was used as an endogenous control. Relative MSI1mRNA levels were plotted against time after actinomycin D addition andfitted to a linear regression model. MSI1 mRNA half-life was calculatedusing the linear regression model. Data were analyzed with the Student ttest and are presented as themean�SEM. Experimentwas carried out intriplicate.

Vo et al.




unaffected byHuR silencing (Fig. 4C), whereas we observeda shift in MSI1 mRNA from heavier polysomal fractions tothe lighter fractions when comparing control cells withHuR-knockdown cells (Fig. 4D). This result indicates thatHuRpromotesMsi1 expression by a combination of increas-ing MSI1 mRNA stability and enhancing its translation.

HuR and Msi1 expression are positively correlatedIt has been suggested thatMsi1 is required tomaintain the

"cancer stem cell population" (18, 40). In support of thisidea, previous studies done in our laboratory and others (30)showed that tumor cells maintained as spheroids, a conditionthat favors the growth of cells with "stem-like" character-istics, have increased Msi1 expression compared with cellsgrown as adherent monolayers. We tested to see whether thesame applies to HuR. HuR expression was measured byqRT-PCR in 2 tumor lines established atUniversity of TexasHealth Science Center at San Antonio revealing also that aswith Msi1, HuR expression levels are higher in spheroidcultures relative to monolayers (Fig. 5A).We then analyzed the mRNA expression of HUR and

MSI1 in clinical samples of glioblastoma by qRT-PCR.When analyzed as on a scatter plot, a statistically significant(P¼ 0.0201) positive correlation ofHUR andMSI1mRNAexpression is observed [Pearson correlation score ¼ 0.503,95% confidence interval (CI), 0.09–0.77], indicating that

glioblastoma tumors that display high Msi1 expression alsoexpress high levels of HuR (Fig. 5B).We extended our expression analysis to examine the

protein expression of both HuR and Msi1. We immunos-tained with both anti-Msi1 and anti-HuR antibodies a tissuemicroarray containing 35 high-quality glioblastoma coresand 5 normal brain tissue cores in duplicate (Fig. 5C). Weobserved that although ubiquitously expressed, HuR is moreabundant in glioblastoma than in normal brain tissue (Fig.5D, tumor average score ¼ 83.48, SEM ¼ 1.507, 95% CI,80.47–86.50 vs. normal tissue average score ¼ 61.00, SEM¼ 9.452, 95% CI, 39.62–82.38, P < 0.0001). As for Msi1,whereas no expression was detected in normal brain, which isconsistent with the notion that Msi1 is not present indifferentiated cells (3), Msi1 expression was often detectedin glioblastoma samples (Fig. 5D and E, tumor average score¼ 41.14, SEM ¼ 4.017, 95% CI, 33.11–49.16 vs. normaltissue average score¼ 0.0, SEM¼ 0.0, 95%CI, 0.0–0.0,P<0.0001). From this analysis, we also ascertain that bothMsi1and HuR expression positively correlate in glioblastoma(Pearson correlation ¼ 0.497, P < 0.0001).

Ectopic Msi1 expression rescues cell proliferationinhibition by HuR silencingWe sought to determine whether Msi1 is a main medi-

ator of HuR function in glioblastoma cells. We silenced

Figure 4. HuR promotes translation ofthe MSI1 mRNA. Sucrose gradientswere used to fractionate lysates fromcontrol U251 cells and cells with HuRknockdown. After fractionation of thesucrose gradient post-ultracentrifugation, RNA from eachfraction was extracted with TRIzol andanalyzed by real-time PCR. GAPDHmRNA, a transcript that is not regulatedby HuR, was included as a negativecontrol. A and B, the results of the A254

during fractionation. The first 3 peaksfrom the left are the 40S, 60S, and 80Ssubunits,whereas thepeaks to the rightare the heavier polysomal fractions.C and D, the percentage distribution ofthe GAPDH (C) or MSI1 (D) is shown.Upon HuR silencing, the distribution ofMSI1 mRNA shifts from the heavierpolysomes to the lighter polysomalfractions while GAPDH mRNA isunaffected. The results of thisexperiment suggest thatHuRpromotesthe translation of the MSI1 mRNA.

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5 6 7 8 9 10 11 12

% G

AP

DH

mR

NA

Fractions

Control siRNA

HuR siRNA

0

5

10

15

20

25

30

0 1 2 3 4 5 6 7 8 9 10 11 12

% M

SI1

mR

NA

Fractions

Control siRNA

HuR siRNA

A B

C D


1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12Fraction number N

o rib

oso

me

co

mp

onents

Rib

osom

e

subun

its

Mo

no

som

e

Lo

w-

mo

lecula

r- w

eig

ht

po

lyso

mes

Hig

h-m

ole

cula

r- w

eig

ht

po

lyso

mes

No

riboso

me

co

mp

onents

Rib

osom

e

subun

its

Mo

no

som

e

Lo

w-

mo

lecula

r- w

eig

ht

po

lyso

mes

Hig

h-m

ole

cula

r- w

eig

ht

po

lyso

mes





0

0.5

1

1.5

2

2.5

3

3.5

082208 monolayer

082208 tumorsphere

092208 monolayer

092208 tumorsphere

Re

lative

mR

NA

exp

ressio

n

MSI1 HuR

A

P = 0.002

P = 0.007P = 1 x 10-6

P = 1 x 10 -4 B

-1

-0.5

0

0.5

1

1.5

2

2.5

-4 -2 0 2 4 6 8 10

ΔΔC

t H

uR

ΔΔ

Correlation = 0.503P = 0.0201n = 21

C

No

rma

l ce

reb

rum

Glio

bla

sto

ma

Msi1 HuRD

P < 0.0001

P < 0.0001

0

10

20

30

40

50

60

70

80

90

Msi1 HuR

Sta

inin

g s

co

re

Glioblastoma tissue, n = 35 Normal brain, n = 5

EP < 0.0001

P < 0.0001

Glio

blas

tom

a tis

sue:

Msi1

Glio

blas

tom

a tis

sue:

HuR

Nor

mal b

rain: M

si1

Nor

mal: H

uR

0

20

40

60

80

100

Sta

inin

g s

core

Ct Msi1

Figure 5. HuR and Msi1 expression levels correlate positively. A,MSI1mRNA and HURmRNA levels are higher in tumor cells with "stem cells" characteristics.HuRandMsi1expressionwereassessed in2primaryglioblastomasamplesgrownasspheroids versusmonolayers. TheACTBmRNAwasused for normalizationwhen using the 2 �DDCtð Þ method for mRNA relative quantification. Data were analyzed with the Student t test and were presented as the mean � SEM.Experimentwascarried out in triplicate. B,HUR andMSI1mRNAexpression positively correlates inGBMs. qRT-PCRwascarried out on total RNAextracted fromthe clinical samples and thedatawere analyzedusing the 2 �DDCtð Þ method. TheACTBmRNAwasusedasanendogenouscontrol and all sampleswere comparedwith the U251 glioblastoma cell line. The DDCtMsi1

was plotted on the abscissa, and the DDCtHuRwas plotted on the ordinate. A linear regression model was

generated with a correlation score of 0.503 and P value of 0.0201; thus, HuR and Msi1 exhibit statistically significant linear correlation in expression.C, representative microscopic images (40�) of immunohistochemistry of Msi1 and HuR in a glioblastoma tissue microarray. D, quantifying glioblastoma tissuemicroarraystainingsofHuRandMsi1.Staining scores forHuRandMsi1 inglioblastoma (35cores)wasaveragedandcomparedwith theaveragescore for staininginnormalbrain tissue (5cores).BothHuRandMsi1aremoreexpressed inglioblastoma than innormalbrain tissue (P< 0.0001).Datawereanalyzedby theStudentt test and presented as the mean� SEM. E, individual scores of each glioblastoma and normal brain tissue samples for HuR and Msi1 are graphed on a verticalscatter plot. Inglioblastoma,HuRhadanaveragescoreof83.48,SEM¼1.507witha95%CI ranging from80.47 to86.50,whereas innormalbrain tissue,HuRhadan average score of 61.00, SEM ¼ 9.452 with a 95% CI from 39.62 to 82.38. In glioblastoma, Msi1 had an average score of 41.14, SEM ¼ 4.017 with a 95% CIranging from 33.11 to 49.16, whereas in normal brain tissue, Msi1 had an average score of 0.0, SEM ¼ 0.0 with a 95% CI ranging from 0.0 to 0.0.

Vo et al.




HuR via siRNA in control lines and in lines contain-ing a MSI1 transgene that does not contain the 30-UTR,rendering the ectopic Msi1 expression immune to anyposttranscriptional regulation. 48 hours after transfection,we measured cell proliferation by MTS assay. While HuRknockdown caused a reduction in cell proliferation incontrol cells (31% reduction in U251 malignant gliomacells and 21% reduction in U343 malignant gliomacells, Fig. 6A and B), the effect was partially amelioratedin cells with transgenic Msi1 expression (13% reduction inU251 and 2.7% in U343, Fig. 6A and B). This resultsuggests that the impact of HuR on cell proliferation ispartially mediated by Msi1.

Ectopic Msi1 expression attenuates the effect of HuRsilencing on cell-cycle distributionWe further determined whether the impact HuR has on

the cell cycle is partially mediated by Msi1. Experimentswere carried out with the U251 and U343 Msi1 and GFPcontrol transgenic cell lines described above. Depletion ofHuR by siRNAs caused an impact on cell-cycle distribu-

tion in both U251 and U343 cells. In U251 cells, HuRsilencing promoted a distribution of cells moving fromG1 to S-phase in U251:GFP cells (Fig. 7A), whereas inU343 cells, we observed an expansion of the population inG2 phase with a concomitant decrease of cells in G1 and aslight decrease in cells found in S-phase (Fig. 7B). WhenHuR silencing was conducted in U251 and U343 cellswhich ectopically express Msi1, the magnitude of theeffect of HuR siRNA on the distribution of cells in eachcell-cycle phase is blunted (Fig. 7A and B). In conclusion,our results indicate that HuR has a profound effect on cell-cycle distribution and suggests that this effect is partiallyachieved via Msi1.

Ectopic Msi1 expression rescues apoptosis induced byHuR silencingBecause HuR and Msi1 have been implicated in the

promotion of a prosurvival/antiapoptotic program in thecell (18, 19, 29, 36, 41–43), we sought to determinewhetherHuR and Msi1 cooperatively function to promote an anti-apoptotic phenotype. When HuR expression is depleted bysiRNA transfection, caspase-3/7 activity is increased 3.3-foldin U251:GFP control cells and 3.9-fold in U343:GFPcontrol cells, Fig. 8A and B, as compared with the controlsiRNA transfection. However, when HuR is depleted ineither the U251:MSI1 or U343:MSI1 stable cell lines,caspase-3/7 activity is only induced 2.3-fold in U251:MSI1cells and 3.2-fold in U343:MSI1 (Fig. 8A and B), suggestingthat Msi1 transgenic expression curtails the induction ofapoptosis when HuR is silenced. This result indicates thatthe antiapoptotic phenotype triggered by HuR is partiallymediated by Msi1.

Discussion

In this study, we explored the posttranscriptional regula-tion of the Msi1 expression by HuR. We found that HuRbinds to sites located in the distal portion of the MSI1 30-UTR and controls its expression viamRNA stabilization andtranslational upregulation. We determined that HuR andMsi1 expression levels are positively correlated and that theidentified regulation has potentially important implicationsfor glioblastoma cell growth.RBPs regulate mammalian gene expression at the post-

transcriptional level, mediating many processes such asmRNA maturation, splicing, transport, storage, stability,and translation. TTR-RBPs (44) are a large family of trans-acting factors that control mRNA stability and translationthrough binding to cis-regulatory elements encoded by anmRNA. Many TTR-RBPs are present in the cell and likelyfunction jointly on a particular mRNA, through additive,antagonistic, or synergistic actions. The role of pathologicand aberrant actions of RBPs in many disease processes,including cancer, has been revealed in recent years. OneRBPof particular interest in cancer is HuR (HuA), a member ofthe Hu/ELAV family of RBPs that also encompasses theneuronal proteins HuB, HuC, and HuD. Through inter-actions with specific subsets of mRNAs, HuR can potentiate

P = 0.003

13%

31%

P = 0.008

2.7%

21%

A

B

0

0.2

0.4

0.6

0.8

1

1.2

U251:GFP U251:MSI1

Re

lative

gro

wth


0

0.2

0.4

0.6

0.8

1

1.2

U343:GFP U343:MSI1

Re

lative

gro

wth


Figure 6. Ectopic expression of Msi1 partially reverses cell proliferationsuppression induced by HuR silencing. A, U251:GFP control and U251:MSI1 cells and (B) U343:GFP control and U343:MSI1 cells weretransfected with either HuR siRNA or scrambled control siRNA. Cellgrowth was measured by MTS assay 48 hours posttransfection. In bothcases, HuR silencing inhibited cell proliferation. However, this effect waspartially recovered when Msi1 was ectopically expressed. Data wereanalyzed by the Student t test and presented as the mean � SEM.Experiment was carried out in triplicate.





biological features of cancers such as sustained proliferation,evasion of apoptosis, increased angiogenesis, and invasion/metastasis (24).HuR is a ubiquitous, multifaceted RBP implicated in

mRNA splicing, translation regulation, and mRNA stabili-zation (24). In recent years, HuR has been implicated as apotent factor in tumorigenesis (24) through its ability to

influences many characteristics of cancer cells (45, 46)including limitless proliferation, sustained angiogenesis,tissue invasion, and metastasis, via its interactions withcancer-related genes [e.g., c-Myc (47), p27 (48), VEGF(49), and Snail (50)]. In high-grade gliomas, prominentHuR expression has been shown; specific mRNA ligands,such as VEGF, COX2, IL-6, IL-8, TGF-b, and TNF-amRNAs, have been identified (25, 26, 28), but themolecularcharacterization of the interaction of HuRwith these ligandshas yet to be carried out. These examples of posttranscrip-tional regulation by HuR suggest that glioma cell growthnecessitates the use of HuR to maintain and augment thegrowth of glioma cells (26).Our study explores a more global effect by HuR in

glioblastoma, through HuR's positive regulation of anotheroncogenic RBP,Msi1, previously implicated in controlling awide subset of mRNAs with cancer-related functions inproliferation, differentiation, survival, and apoptosis (51).Our findings posit that HuR not only elicits direct effects oncancer cell gene expression but also can exert indirect gene-regulatory effects by controlling expression of another keyregulator of gene expression, Msi1. Our findings may beextended to other malignancies that depend on Msi1

A

B

P = 0.0004

P = 0.0002

P = 0.005

P = 0.0067

P = 0.0067

-20

-15

-10

-5

0

5

10

15

20

U251:GFP:HuR

siRNA vs. control

siRNA

U251:MSI1:HuR

siRNA vs. control

siRNA

Diffe

rence o

f perc

enta

ge in e

ach

phase (

com

pare

d w

ith c

ontr

ol

siR

NA

)

G1

S

G2

-6

-4

-2

0

2

4

6

U343:GFP:HuR

siRNA vs. control

siRNA

U343:MSI1:HuR

siRNA vs. control

siRNA

Diffe

rence o

f perc

enta

ge in e

ach

phase (

com

pare

d w

ith c

ontr

ol

siR

NA

)

G1

S

G2

Figure 7. Cell-cycle distribution is altered upon HuR silencing withsiRNAs and is partially attenuated with ectopic Msi1 expression. Cell-cycle analysis was conducted using propidium iodide staining ofethanol-fixed cells for DNA content which was analyzed using flowcytometry. A and B, U251:GFP control and U251:MSI1 cells weretransfected with either HuR siRNA or scrambled control siRNA. Asimilar experiment was carried out in U343:GFP and U343:MSI1 cells.Cell-cycle distribution was analyzed using the Modfit LT SoftwarePackage. Then, the percentage of cells in each phase found in the HuRsiRNA experiment was subtracted from the control siRNA experiment.HuR silencing caused a large shift in cell-cycle distribution in theU251:GFP (A) and U343:GFP (B) cells. However, when Msi1 isectopically expressed, the change in cell-cycle distribution induced byHuR silencing was attenuated (A and B). Data were analyzed by theStudent t test and presented as the mean � SEM. Experiment wascarried out in triplicate.

P = 0.004

P = 0.009A

B

0

0.5

1

1.5

2

2.5

3

3.5

4

U251:GFP U251:MSI1

Fo

ld c

asp

ase-3

/7 a

ctiva

tio

n


0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

U343:GFP U343:MSI1

Fo

ld c

asp

ase-3

/7 a

ctiva

tio

n


Δ = 0.96- fold

Δ = 0.75- fold

Figure 8. Ectopic expression of Msi1 partially recovers the effect of HuRsilencing on apoptosis. A, HuR or control siRNAs were transfected inU251:GFP and U251:MSI1 cells. B, similar experiments were carried outin U343:GFP and U343:MSI1 cells. Apoptosis was measured 48 hoursposttransfection using a luciferase-based reagent to monitor caspase-3/7 activity. HuR silencing was able to induce apoptosis, whereas Msi1ectopic expression partially inhibits this effect. Datawere analyzed by theStudent t test and presented as themean�SEM. Experimentwas carriedout in triplicate.

Vo et al.




expression such as medulloblastoma, hepatocellular carci-noma, cervical adenocarcinoma, non–small cell lung cancer,breast and colon cancer (8–10, 18, 19). High HuR expres-sion has also been reported in other studies (25, 31, 52–54);the overlap in expression profiles of Msi1 and HuR suggeststhat the HuR:Msi1 relationship described in this report mayexist in other tumors (Table 1). From these results, acomplex posttranscriptional genetic regulatory networkemerges (55) for glioblastoma that will describe the cellularand physiologic functions required for gliomagenesisthrough the discovery of network motifs (56), modularbuilding blocks of gene regulatory networks.The HuR:Msi1 relationship discussed in this article cre-

ates a positive cascade network motif (56) where the cellularlogic for gliomagenesis requires HuR to essentially "activate"Msi1 gene expression. Moreover, HuR also autoregulatesitself bolstering its own production in a positive feedbackloop (57–59). The interaction between these 2 RBPs createsa 2-component composite gene regulatory network. InSaccharomyces cerevisiae, 21 of the 46 studied RBPs arethought to be involved in a two- or multicomponentcomposite gene regulatory network, thus revealing thepotential for a dense genetic regulatory network that hasnot been previously recognized (55).

However, posttranscriptional genetic regulatory networkmotifs are not limited by RBP nodes but extend to micro-RNAs. A recent study from our laboratory illustrates theposttranscriptional regulation of Musashi1 by tumor sup-pressor miRNAs (22). Moreover, we also recently showedthat 25% of HuR-binding sites in 30-UTRs mapped byCLIP (cross-linking and immunoprecipitation) coincidewith miRNA predicted sites, suggesting an interplaybetween RBPs and miRNAs (60). These complex relation-ships show the nonlinearity of posttranscriptional networkentities and shows that their integration creates a larger,complex network that contains the pertinent logic for themolecular pathogenesis in glioblastoma. Furthermore, post-transcriptional gene networks are intimately intertwinedwith the transcriptional gene network thus an inclusiveanalysis of the circuitry required for glioblastoma tumori-genesis surfaces, providing insights into glioblastoma cellularlogic and physiology and the emergence of genetic network–directed therapies (61–63).

Disclosure of Potential Conflict of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Suzanne C. Burns for critical reading and comments on themanuscript and the anonymous referees for their critical reading and suggestions for themanuscript throughout the review process.

Grant Support

K. Abdelmohsen, J.L.Martindale, K. Tominaga, andM.Gorospe are supported bythe Intramural Research Program of the National Institute on Aging, NIH. V. Patel issupported by the Intramural Research Program at the National Institute of Dental andCraniofacial Research, NIH. D.T. Vo, M. Qiao, T.L. Burton, and L.O.F. Penalva aresupported by grants from the Children's Brain Tumor Foundation, Association forResearch ofChildhoodCancer, andTheMax andMinnie Tomerlin Voelcker Fund. A.J. Brenner is supported by the Cancer Therapy &Research Center P30 Cancer CenterSupport Grant from the National Cancer Institute (CA054174).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be herebymarked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Received May 8, 2011; revised October 9, 2011; accepted October 28, 2011;published online January 18, 2012.

References1. Glisovic T, Bachorik JL, Yong J, Dreyfuss G. RNA-binding proteins

and post-transcriptional gene regulation. FEBS Lett 2008;582:1977–86.

2. Nakamura M, Okano H, Blendy JA, Montell C. Musashi, a neural RNA-binding protein required for Drosophila adult external sensory organdevelopment. Neuron 1994;13:67–81.

3. Sakakibara S, Imai T, Hamaguchi K, Okabe M, Aruga J, Nakajima K,et al. Mouse-Musashi-1, a neural RNA-binding protein highly enrichedin the mammalian CNS stem cell. Dev Biol 1996;176:230–42.

4. Kayahara T, SawadaM,Takaishi S, FukuiH, SenoH, FukuzawaH, et al.Candidate markers for stem and early progenitor cells, Musashi-1 andHes1, are expressed in crypt base columnar cells of mouse smallintestine. FEBS Lett 2003;535:131–5.

5. Clarke RB. Isolation and characterization of human mammary stemcells. Cell Prolif 2005;38:375–86.

6. Sugiyama-Nakagiri Y, Akiyama M, Shibata S, Okano H, Shimizu H.Expression of RNA-binding protein Musashi in hair follicle devel-opment and hair cycle progression. Am J Pathol 2006;168:80–92.

7. Yokota N, Mainprize TG, Taylor MD, Kohata T, Loreto M, Ueda S, et al.Identification of differentially expressed and developmentally regulat-ed genes in medulloblastoma using suppression subtraction hybrid-ization. Oncogene 2004;23:3444–53.

8. Shu HJ, Saito T, Watanabe H, Ito JI, Takeda H, Okano H, et al.Expression of the Musashi1 gene encoding the RNA-binding proteinin human hepatoma cell lines. Biochem Biophys Res Commun2002;293:150–4.

9. Ye F, Zhou C, Cheng Q, Shen J, Chen H. Stem-cell-abundant proteinsNanog,Nucleostemin andMusashi1 are highly expressed inmalignantcervical epithelial cells. BMC Cancer 2008;8:108.

Table 1. Summary of malignancies that exhibitincreased HuR and Msi1 expression

Tumor type HuR Msi1 References

Glioblastoma " " (12, 25, 64)Medulloblastoma " " (18, 25)Hepatocellularcarcinoma

" " (8, 31)

Cervical adenocarcinoma " " (9, 52)Non–small celllung cancer

" " (10, 53)

Colon cancer " " (19, 59)

NOTE: " indicates increased expression. References foreach study are called out in the rightmost column.





10. Kanai R, Eguchi K, TakahashiM,GoldmanS,OkanoH,KawaseT, et al.Enhanced therapeutic efficacy of oncolytic herpes vector G207against human non-small cell lung cancer–expression of an RNA-binding protein, Musashi1, as a marker for the tailored gene therapy.J Gene Med 2006;8:1329–40.

11. Haigis KM, Kendall KR, Wang Y, Cheung A, Haigis MC, Glickman JN,et al. Differential effects of oncogenic K-Ras and N-Ras on prolifer-ation, differentiation and tumor progression in the colon. Nat Genet2008;40:600–8.

12. Kanemura Y, Mori K, Sakakibara S, Fujikawa H, Hayashi H, Nakano A,et al. Musashi1, an evolutionarily conserved neural RNA-bindingprotein, is a versatile marker of human glioma cells in determiningtheir cellular origin, malignancy, and proliferative activity. Differentia-tion 2001;68:141–52.

13. Wang XY, Penalva LO, Yuan H, Linnoila RI, Lu J, Okano H, et al.Musashi1 regulates breast tumor cell proliferation and is a prognosticindicator of poor survival. Mol Cancer 2010;9:221.

14. Glazer RI, Wang XY, Yuan H, Yin Y. Musashi1: a stem cell marker nolonger in search of a function. Cell Cycle 2008;7:2635–9.

15. Guessous F, Li Y, Abounader R. Signaling pathways in medulloblas-toma. J Cell Physiol 2008;217:577–83.

16. Lino MM, Merlo A, Boulay JL. Notch signaling in glioblastoma: adevelopmental drug target? BMC Med 2010;8:72.

17. Imai T, Tokunaga A, Yoshida T, Hashimoto M, Mikoshiba K, Wein-master G, et al. The neural RNA-binding protein Musashi1 translation-ally regulatesmammalian numbgene expression by interactingwith itsmRNA. Mol Cell Biol 2001;21:3888–900.

18. Sanchez-Diaz PC, Burton TL, Burns SC, Hung JY, Penalva LO.Musashi1 modulates cell proliferation genes in the medulloblastomacell line Daoy. BMC Cancer 2008;8:280.

19. Sureban SM, May R, George RJ, Dieckgraefe BK, McLeod HL,Ramalingam S, et al. Knockdown of RNA binding protein musa-shi-1 leads to tumor regression in vivo. Gastroenterology 2008;134:1448–58.

20. Ratti A, Fallini C, Cova L, Fantozzi R, Calzarossa C, Zennaro E, et al. Arole for the ELAV RNA-binding proteins in neural stem cells: stabili-zation of Msi1 mRNA. J Cell Sci 2006;119:1442–52.

21. Okano H. Adult neural stem cells and central nervous system repair.Ernst Schering Res Found Workshop 2006:215–28.

22. Vo DT, Qiao M, Smith AD, Burns SC, Brenner AJ, Penalva LO. Theoncogenic RNA-binding protein Musashi1 is regulated by tumor sup-pressor miRNAs. RNA Biol 2011 Sep 1;8(5).

23. Barreau C, Paillard L, Osborne HB. AU-rich elements and associatedfactors: are there unifying principles? Nucleic Acids Res 2005;33:7138–50.

24. Abdelmohsen K, Gorospe M. Posttranscriptional regulation of cancertraits by HuR. Wiley Interdiscip Rev–RNA 2010;1:214–29.

25. Nabors LB, Gillespie GY, Harkins L, King PH. HuR, a RNA stabilityfactor, is expressed in malignant brain tumors and binds to adenine-and uridine-rich elements within the 30 untranslated regions ofcytokine and angiogenic factor mRNAs. Cancer Res 2001;61:2154–61.

26. Bolognani F, Gallani AI, Sokol L, Baskin DS, Meisner-Kober N. mRNAstability alterations mediated by HuR are necessary to sustain the fastgrowth of glioma cells. J Neurooncol 2011 Sep 21. [Epub ahead ofprint].

27. Brennan CM, Steitz JA. HuR and mRNA stability. Cell Mol Life Sci2001;58:266–77.

28. NaborsLB, SuswamE,HuangY,YangX, JohnsonMJ,KingPH. Tumornecrosis factor alpha induces angiogenic factor up-regulation inmalignant glioma cells: a role for RNA stabilization and HuR. CancerRes 2003;63:4181–7.

29. Filippova N, Yang X, Wang Y, Gillespie GY, Langford C, King PH, et al.The RNA-binding protein HuR promotes glioma growth and treatmentresistance. Mol Cancer Res 2011;9:648–59.

30. Yuan X, Curtin J, Xiong Y, Liu G, Waschsmann-Hogiu S, Farkas DL,et al. Isolation of cancer stem cells from adult glioblastomamultiforme.Oncogene 2004;23:9392–400.

31. Mazan-Mamczarz K, Galb�an S, L�opez de Silanes I, Martindale JL,Atasoy U, Keene JD, et al. RNA-binding protein HuR enhances p53

translation in response to ultraviolet light irradiation. ProcNatl AcadSciU S A 2003;100:8354–9.

32. GlasM, Rath BH, SimonM, Reinartz R, Schramme A, Trageser D, et al.Residual tumor cells are unique cellular targets in glioblastoma. AnnNeurol 2010;68:264–9.

33. Lee J, KotliarovaS, Kotliarov Y, Li A, SuQ,DoninNM, et al. Tumor stemcells derived from glioblastomas cultured in bFGF and EGF moreclosely mirror the phenotype and genotype of primary tumors thando serum-cultured cell lines. Cancer Cell 2006;9:391–403.

34. AmornphimolthamP, Patel V, Sodhi A,NikitakisNG, Sauk JJ, SausvilleEA, et al. Mammalian target of rapamycin, a molecular target insquamous cell carcinomas of the head and neck. Cancer Res2005;65:9953–61.

35. Lopez de Silanes I, ZhanM, Lal A, Yang X, GorospeM. Identification ofa target RNAmotif for RNA-binding protein HuR. Proc Natl Acad Sci US A 2004;101:2987–92.

36. Lal A, Kawai T, YangX,Mazan-MamczarzK,GorospeM.Antiapoptoticfunction of RNA-binding protein HuR effected through prothymosinalpha. EMBO J 2005;24:1852–62.

37. Belizario JE, Alves J, Garay-Malpartida M, Occhiucci JM. Couplingcaspase cleavage and proteasomal degradation of proteins carryingPEST motif. Curr Protein Pept Sci 2008;9:210–20.

38. Ikeda E, Achen MG, Breier G, Risau W. Hypoxia-induced transcrip-tional activation and increased mRNA stability of vascular endo-thelial growth factor in C6 glioma cells. J Biol Chem 1995;270:19761–6.

39. Perry RP, Kelley DE. Messenger RNA-protein complexes and newlysynthesized ribosomal subunits: analysis of free particles and com-ponents of polyribosomes. J Mol Biol 1968;35:37–59.

40. Wang XY, Yin Y, Yuan H, Sakamaki T, Okano H, Glazer RI. Musashi1modulates mammary progenitor cell expansion through proliferin-mediated activation of the Wnt and Notch pathways. Mol Cell Biol2008;28:3589–99.

41. Ishimaru D, Ramalingam S, Sengupta TK, Bandyopadhyay S, Dellis S,TholanikunnelBG, et al. RegulationofBcl-2 expressionbyHuR inHL60leukemia cells and A431 carcinoma cells. Mol Cancer Res 2009;7:1354–66.

42. Durie D, Lewis SM, Liwak U, Kisilewicz M, GorospeM, Holcik M. RNA-binding protein HuR mediates cytoprotection through stimulation ofXIAP translation. Oncogene 2011;30:1460–9.

43. Abdelmohsen K, Lal A, Kim HH, Gorospe M. Posttranscriptionalorchestration of an anti-apoptotic program by HuR. Cell Cycle 2007;6:1288–92.

44. Abdelmohsen K, Kuwano Y, Kim HH, Gorospe M. Posttranscrip-tional gene regulation by RNA-binding proteins during oxidativestress: implications for cellular senescence. Biol Chem 2008;389:243–55.

45. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57–70.

46. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation.Cell 2011;144:646–74.

47. Kim HH, Kuwano Y, Srikantan S, Lee EK, Martindale JL, Gorospe M.HuR recruits let-7/RISC to repress c-Myc expression. Genes Dev2009;23:1743–8.

48. Kullmann M, Gopfert U, Siewe B, Hengst L. ELAV/Hu proteins inhibitp27 translation via an IRES element in the p27 50UTR. Genes Dev2002;16:3087–99.

49. Levy NS, Chung S, Furneaux H, Levy AP. Hypoxic stabilization ofvascular endothelial growth factor mRNA by the RNA-binding proteinHuR. J Biol Chem 1998;273:6417–23.

50. Dong R, Lu JG, Wang Q, He XL, Chu YK, Ma QJ. Stabilization of Snailby HuR in the process of hydrogen peroxide induced cell migration.Biochem Biophys Res Commun 2007;356:318–21.

51. deSousa AbreuR, Sanchez-Diaz PC, Vogel C, Burns SC, KoD, BurtonTL, et al. Genomic analyses of musashi1 downstream targets show astrong association with cancer-related processes. J Biol Chem2009;284:12125–35.

52. Kawai T, Lal A, Yang X, Galban S, Mazan-Mamczarz K, Gorospe M.Translational control of cytochrome c by RNA-binding proteins TIA-1and HuR. Mol Cell Biol 2006;26:3295–307.

Vo et al.




53. Wang J, ZhaoW, Guo Y, Zhang B, Xie Q, Xiang D, et al. The expressionof RNA-binding protein HuR in non-small cell lung cancer correlateswith vascular endothelial growth factor-C expression and lymph nodemetastasis. Oncology 2009;76:420–9.

54. Lal A, Mazan-Mamczarz K, Kawai T, Yang X, Martindale JL,Gorospe M. Concurrent versus individual binding of HuR andAUF1 to common labile target mRNAs. EMBO J 2004;23:3092–102.

55. Kanitz A, Gerber AP. Circuitry of mRNA regulation. Wiley InterdiscipRev Syst Biol Med 2010;2:245–51.

56. Alon U. Networkmotifs: theory and experimental approaches. Nat RevGenet 2007;8:450–61.

57. Pullmann R, Kim HH, Abdelmohsen K, Lal A, Martindale JL, Yang X,et al. Analysis of turnover and translation regulatory RNA-bindingprotein expression through binding to cognate mRNAs. Mol Cell Biol2007;27:6265–78.

58. Yi J, Chang N, Liu X, Guo G, Xue L, Tong T, et al. Reduced nuclearexport of HuR mRNA by HuR is linked to the loss of HuR in replicativesenescence. Nucleic Acids Res 2010;38:1547–58.

59. Al-AhmadiW,Al-GhamdiM,Al-Haj L, Al-SaifM,KhabarKS. Alternativepolyadenylation variants of the RNA binding protein, HuR: abundance,role of AU-rich elements and auto-Regulation. Nucleic Acids Res2009;37:3612–24.

60. Uren PJ, Burns SC, Ruan J, Singh KK, Smith AD, Penalva LO.Genomic analyses of the RNA binding protein Hu Antigen R(HuR) identify a complex network of target genes and novelcharacteristics of its binding sites. J Biol Chem 2011;286:37063–6.

61. MorrowJK, Tian L, ZhangS.Molecular networks in drugdiscovery. CritRev Biomed Eng 2010;38:143–56.

62. Kitano H. Computational systems biology. Nature 2002;420:206–10.

63. Kitano H. A robustness-based approach to systems-oriented drugdesign. Nat Rev Drug Discov 2007;6:202–10.

64. TodaM, Iizuka Y, YuW, Imai T, Ikeda E, Yoshida K, et al. Expression ofthe neural RNA-binding protein Musashi1 in human gliomas. Glia2001;34:1–7.





2012;10:143-155. Mol Cancer Res Dat T. Vo, Kotb Abdelmohsen, Jennifer L. Martindale, et al. HuR via mRNA Translation and Stability in Glioblastoma CellsThe Oncogenic RNA-Binding Protein Musashi1 Is Regulated by

Updated version

http://mcr.aacrjournals.org/content/10/1/143

Access the most recent version of this article at:

Cited articles

http://mcr.aacrjournals.org/content/10/1/143.full#ref-list-1

This article cites 61 articles, 20 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/10/1/143.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mcr.aacrjournals.org/content/10/1/143To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/content/10/1/143.full#ref-list-1

http://mcr.aacrjournals.org/content/10/1/143.full#related-urls

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

mailto:[email protected]



Documents

TheOncogenicRNA-BindingProteinMusashi1IsRegulatedby HuR via mRNA Translation … · Musashi1(Msi1)isanevolutionarilyconservedRNA-bindingprotein(RBP)thathasprofoundimplicationsin cellular