o-glcnac transferase integrates metabolic pathways to ... · o-glcnac transferase integrates...

Tumor and Stem Cell Biology

O-GlcNAc Transferase Integrates Metabolic Pathways toRegulate the Stability of c-MYC in Human Prostate CancerCells

Harri M. Itkonen1, Sarah Minner5, Ingrid J. Guldvik1, Mareike Julia Sandmann5, Maria Christina Tsourlakis5,Viktor Berge3, Aud Svindland4, Thorsten Schlomm6,7, and Ian G. Mills1,2,3

AbstractMetabolic disruptions that occur widely in cancers offer an attractive focus for generalized treatment strategies.

The hexosamine biosynthetic pathway (HBP) senses metabolic status and produces an essential substrate forO-linked b-N-acetylglucosamine transferase (OGT), which glycosylates and thereby modulates the function of itstarget proteins. Here,we report that theHBP is activated inprostate cancer cells and thatOGT is a central regulatorof c-Myc stability in this setting.HBP geneswere overexpressed in human prostate cancers and androgen regulatedin cultured human cancer cell lines. Immunohistochemical analysis of human specimens (n ¼ 1987) establishedthat OGT is upregulated at the protein level and that its expression correlates with high Gleason score, pT and pNstages, and biochemical recurrence. RNA interference–mediated siliencing or pharmacologic inhibition of OGTwas sufficient todecreaseprostate cancer cell growth.Microarrayprofiling showed that theprincipal effectsofOGTinhibition in prostate cancer cells were related to cell-cycle progression and DNA replication. In particular, c-MYCwas identified as a candidateupstreamregulator ofOGT target genes andOGT inhibition elicited a dose-dependentdecrease in the levels of c-MYC protein but not c-MYCmRNA in cell lines. Supporting this relationship, expressionof c-MYC andOGTwas tightly correlated in human prostate cancer samples (n¼ 1306). Our findings identify HBPas amodulator of prostate cancer growth and c-MYCas a key target ofOGT function inprostate cancer cells.CancerRes; 73(16); 5277–87. �2013 AACR.

IntroductionProstate cancer is the second most common male cancer in

the world. The androgen receptor (AR) is a principal targetin prostate cancer research because AR activity is maintainedin castration-resistant disease, and both localized andadvanced diseases are responsive to drugs that alter hormonalsignaling (1). AR regulates anabolic metabolism and promotesaerobic glycolysis (2, 3). Experiments carried out in cell lineshave led to the identification of large metabolic networks andthe question remains whether there is an integration point ofthese networks that is clinically relevant.The hexosamine biosynthetic pathway (HBP) requires glu-

tamine, glucose, acetyl-Coenzyme-A, and nucleotide UTP to

synthesize UDP-N-acetyl-D-glucosamine (UDP-GlcNAc; ref. 4).This pathway senses the availability of energy and couplesmetabolic flux to control cell proliferation (4–6). HBP providessubstrate for postranslational modification of plasma mem-brane and secretory proteins. In addition, UDP-GlcNAc is usedby O-linked b-N-acetylglucosamine (O-GlcNAc) transferase(OGT) that modifies target proteins in cytosol, mitochondria,and nucleus (7). Consequently, the HBP has emerged as aversatile regulator of signaling cascades influencing cell cycle(8), growth (5), metabolism (6, 9), and stress (7, 10). OGT haspreviously been reported to be overexpressed in a range ofcancers, such as breast cancer (11) and lung and colon cancers(12). Knockdown of OGT reduces tumor growth in breastcancer mouse model (11) and invasion in colon cancer celllines (12), suggesting that OGT activity contributes to thetransformation phenotype. Some reports suggest that OGTserves similar role in prostate cancer cell lines, in part, throughthe O-linked glycosylation of FOXM1 (13), but no study has yetexplored the regulation of the rate-limiting steps in the HBPpathway or assessed the effect ofOGT inhibition in anunbiasedway.

Materials and MethodsProstate tissue specimens for mRNA analysis

Matched benign and malignant prostate tissues werederived from radical prostatectomy of 29 patients with pros-tate cancer, treated at Oslo Urological University Clinic (Oslo,

Authors' Affiliations: 1Prostate Research Group, Centre for MolecularMedicine Norway, Nordic EMBL Partnership, University of Oslo and OsloUniversity Hospital; 2Departments of Cancer Prevention, 3Urology, and4Pathology, Oslo University Hospital, Oslo, Norway; 5Institute for Pathol-ogy; 6Department of Urology, Section for Translational Prostate CancerResearch; and 7Martini-Clinic, Prostate Cancer Center, University MedicalCenter Hamburg-Eppendorf, Hamburg, Germany

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

Corresponding Author: Ian Geoffrey Mills, Centre for Molecular MedicineNorway (NCMM), P.O. Box 1137, Blindern, 0318 Oslo, Norway. Phone: 47-228-40767; Fax: 47-228-40598; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-0549

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 5277

on July 8, 2020. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 29, 2013; DOI: 10.1158/0008-5472.CAN-13-0549

http://cancerres.aacrjournals.org/

Norway) and obtained from "The ProstateBiobank - a resourcefor urological research in Norway" (http://www.ous-research.no/home/tasken/Research%20interests/10769). Clinical para-meters of this prostate cancer patient cohort are available inSupplementary Table S3. Samples were treated with RNALaterat 4�C overnight according to the manufacture's instructions(Qiagen) and stored at �80�C. The pathology of the sampleswas verified by uropathologist. Written consent was obtainedfrom all patients, and the project was approved by TheRegional Committee for Medical and Health Research Ethics(REK sør-øst 2010/2218). One microgram of total RNA wasreverse transcribed using the High-Capacity RNA-to-cDNAMaster Mix (Applied BioSystems) according to manufacturer'sinstructions. Quantitative real-time PCR (RT-PCR) was con-ducted using 1 ng of cDNA with Fast SYBR Green PCRMasterMix (Applied Biosystems) for the detection of GFPT1,OGT, and UAP1 (primer sequences are given in SupplementaryTable S4). TaqMan Gene expression Assays (Applied Biosys-tems) were used to determine the mRNA expression of the AR,ERG, MYC, GADPH, TBP, ALAS1, b-ACTIN, and 18S. An ABIHT7900 Sequence Detection System was used according tomanufacturer's recommendations. Each cDNA sample wastested in triplicate and mean Ct values were calculated. ThemRNA expression level was determined using the comparativeCt method and normalization to the geometric mean ofGAPDH, ALAS1, 18s, ACTB, and TBP. The expression level ofthe gene of interest in the normal tissue was set to the value of1, and the values for the matched tumor sample were calcu-lated accordingly. The difference inmRNA expression betweenmatched normal and tumor sample were tested by a depen-dent sample t test. All data were analyzed using SPSS v.19software (SPSS).

Tissue microarray and immunohistochemistryPatient cohort. Radical prostatectomy specimens were

available from 3,261 patients, treated in the Department ofUrology, University Medical Center Hamburg-Eppendorf(Hamburg, Germany) between 1992 and 2008 (14). Follow-updata were available for 3,057 patients, ranging from 0.1 to 228.7months (mean 101.6 months). None of the patients receivedneoadjuvant endocrine therapy. Additional (salvage) therapywas initiated in case of a biochemical relapse (BCR). In allpatients, prostate-specific antigen (PSA) valuesweremeasuredquarterly in the first year, followed by biannual measurementsin the second, and annual measurements after the third yearfollowing surgery. Recurrence was defined as a postoperativePSA of 0.2 ng/mL and rising thereafter. The first PSA valueabove or equal to 0.2 ng/mL was used to define the time ofrecurrence. Patients without evidence of tumor recurrencewere censored at the last follow-up. All prostatectomy speci-mens were analyzed according to a standard procedure. Allprostates were completely paraffin embedded, includingwhole-mount sections as previously described (15). All hema-toxylin and eosin stained histologic sections from all prosta-tectomy specimens were reviewed and one 0.6mm thick tissuecore was punched out from a representative cancer area andtransferred onto a tissue microarray (TMA) format asdescribed (16). The 3,261 cores were distributed among 7 TMA

blocks each containing 129 to 522 tumor samples. For internalcontrols, each TMA block also contained different variouscontrol tissues, including normal prostate tissue.

Immunohistochemistry. Following immunohistochemis-try (IHC) optimization for OGT (mouse monoclonal; NovusBiologicals #NB300-524; dilution 1:1350), the TMA was immu-nostained. Slides were deparaffinized and exposed to heat-induced antigen retrieval for 5 minutes in an autoclave at pH9.0. Bound primary antibody was visualized using the DAKOEnVision Kit (DAKO).

Nuclear OGT staining (0, þ1, þ2, þ3) was scored for eachtissue spot. This scoring pattern has been used previously (17).Tissue samples without definite prostate cancer were excluded.

Statistical analysis. Statistical calculations were con-ducted with JMP statistical software (Version 8.0, SAS insti-tute). Contingency tables were calculated with the c2 test andFisher exact test to analyze differences between groups. Sur-vival curves were calculated by the Kaplan–Meier method andcompared with the log-rank test.

Technical issues. As in all TMA studies, a fraction of thecases were noninformative due to complete lack of tissuesamples (n ¼ 475) or absence of unequivocal cancer tissue(n ¼ 799).

Cell lines and maintenanceCells were obtained from the American Tissue Culture

Collection (ATCC) and maintained according to ATCCguidelines. Cell lines were authenticated by the providerand were used within 6 months receipt. LNCaP and PC-3cells were grown in RPMI media supplemented with 10%FBS. VCaP cells were cultured in Dulbecco's Modified EagleMedium supplemented with 10% FBS. RWPE-1 cells weremaintained in Keratinocyte-SFM (KSFM) media supplemen-ted with EGF and bovine pituitary extract, according tomanufacturer's instructions. Before R1881 stimulation, cellswere maintained in phenol red-free media supplementedwith 10% charcoal-stripped serum for 72 hours (LNCaP andVCaP), whereas RWPE-1 cells were maintained in KSFMmedia supplemented with 0.5% bovine serum albumin. Theviability of cells was assessed with an MTS assay (Promega)according to manufacturer's instructions.

Preparation of cell lysates and Western blottingAll the steps were conducted at 4�C, unless otherwise

mentioned. Cells were washed once with PBS and harvestedinto cell lysis buffer (10 mmol/L Tris-HCl, pH 8.0, 1 mmol/LEDTA, 0.5 mmol/L EGTA, 1% Triton X-100, 0.1% sodium-deoxycholate, 0.1% SDS, 140 mmol/L NaCl þ Complete pro-tease inhibitor mixture; Roche), rotated for 15 minutes andcentrifuged 14,000 g 10 minutes. Supernatant was collectedand protein concentration determinedwith bicinchoninic acid(BCA) assay. Of note, 10 to 25 mg of lysate was separated withSDS-PAGE, using 4% to 12% gradient gels (Invitrogen) andtransferred to nitrocellulose membranes. Membranes wereprobed with antibodies against OGT (Santa Cruz), GFPT1,ACTIN, b-tubulin, UAP1 (Sigma), MYC (Epitomic), and RL2(Abcam). Primary antibodies were detected with horseradishperoxidase -conjugated secondary antibodies against cognate

Itkonen et al.

Cancer Res; 73(16) August 15, 2013 Cancer Research5278




species (Dako). The intensity of the signals from each antibodywas quantified by Quantity One software (Bio-Rad).

Immunoprecipitation and lectin pulldownAll the steps were conducted at 4�C. Cells were washed once

with PBS and solubilized in cell lysis buffer (10 mmol/L Tris-HCl, pH 8.0, 1 mmol/L EDTA, 0.5 mmol/L EGTA, 1% Triton X-100, 0.1% Nadeoxycholate, 0.1% SDS, 140 mmol/L NaCl þComplete protease inhibitor mixture; Roche), rotated for 15minutes and centrifuged 18,000 g for 5 minutes. Proteinconcentration was determined with BCA assay and 1,000 to3,000 mg of protein was precleared with unspecific antibody(Santa Cruz) and protein A-goated magnetic beads (immuno-precipitation) or unbound agarose beads (Vector Labs) for 2hours. Precleared extract was used for immunoprecipitation(RL2 antibody, Abcam), or lectin pulldown (Wheat GermAgglutinin, WGA, VectorLabs) overnight. Protein G-goatedmagnetic beads were added to the immunoprecipitation reac-tion, incubated for 2 hours and washed with immunoprecip-itation wash buffer (0.5% NP-40, 150 mmol/L NaCl, 20 mmol/LTris-HCl, pH 8.0). Lectin pulldown was washed 3 times withlectin wash buffer (0.1% Tween, 150 mmol/L NaCl, 10 mmol/LTris-HCl, pH 8.0).

RT-PCR and expression profiling of the cell linesRNA was collected by illustraMiniSpin (GE Healthcare)

according to manufacturer's instructions. One microgram ofRNA was used to produce cDNA (qScript cDNA Synthesis Kit,Quanta Biosciences). Subsequently, 0.3 mL was used for quan-titative PCR (qPCR). Amplification was conducted as follows:10 minutes 95�C followed by 40 cycles 30 seconds 60�C, 30seconds extension, final extension 5 minutes in 72�C. Genesdetected with SYBR Green are listed in the SupplementaryTable S4 with corresponding primers. Housekeeping geneswere detected with TaqMan assays (Applied Biosystems).

MicroarraysThepurity and quantity of the extractedRNAweremeasured

using the NanoDrop ND1000 spectrophotometer (NanodropTechnologies), and the RNA integrity was evaluated using theAgilent 2100 Bioanalyzer with the RNA nano 6000 kit (AgilentTechnologies Inc.,). Synthesis of cDNA, cRNA amplification,and hybridization of cRNA onto the Illumina HumanHT-12 v.4Expression BeadChips (Illumina Inc,) was carried out as permanufacturer's instructions. Data extraction, quantile normal-ization, and initial quality control of the bead summaryraw data were conducted using GenomeStudio v2011.1 fromIllumina and theGene Expressionmodule v1.9.0. The datawereannotated using the HumanHT-12_V4_0_R2_15002873_B.bgxannotation file from Illumina. Microarray data are depositedin Gene Expression Omnibus with an accession numberGSE44624.

TreatmentsSynthetic androgen, R1881 was solubilized in ethanol to a

final concentration of 10 mmol/L. Inhibitor against humanO-GlcNAc transferase, ST045849 (3-(2-adamantanylethyl)-2-[(4-chlorophenyl)azamethylene]-4-oxo-1,3-thiazaperhyd roine-

6-carboxylic acid), was purchased from TimTec and solubilizedin dimethyl sulfoxide to final concentration of 20mmol/L. OGTinhibitor has an IC50 value of 30 to 53mmol/L, depending on theisoformof the enzyme (18) and it has beenused to targetOGT in100 mmol/L (19). siRNAs targeting OGT (siOGT) were obtainedfrom Lifetechnologies (s16094 and s16095), and RNAiMax wasused for the transfection.

ResultsHexosamine biosynthetic pathway is upregulated duringearly stages of prostate cancer

We have previously reported that the AR is a regulator ofboth metabolic and cell-cycle gene networks (2). In particular,we and others have found that genes associatedwithmetabolicpathways are significantly overexpressed in prostate cancer (2,3, 20). By combining publicly available chromatin immuno-precipitation (ChIP) data for 2 key transcription factors inprostate cancer (AR; ref. 2; ERG; ref. 21) and clinical geneexpression data from11 separate studies (20), we identified 195genes, which have binding sites within their promoters and areoverexpressed in the clinical setting (Supplementary Table S1).A pathway analysis on this geneset highlighted O-glycanbiosynthesis and amino and sugar nucleotide metabolism assignificantly deregulated processes (Supplementary Table S1).As a classifier, O-glycan biosynthesis encompasses complexsugar chains attached to plasma membrane proteins (22) andsingle O-GlcNAcmodification occurring in other cell compart-ments (7). Importantly, single O-GlcNAc modification is cat-alyzed by only one enzyme in humans (OGT), which requiresthe activity of HBP and therefore functions as a metabolicintegration point. The flux through HBP is controlled by therate-limiting enzyme glutamine-fructose-6-phosphate trans-aminase 1 (GFPT1).

Biomarker validationWe sought to validate the transcript expression of the HBP

genes in clinical samples. For this, we had access to 29 prostatecancer patient samples (matched tumor and normal tissue).We found that the enzymes catalyzing the rate-limiting step(GFPT1) and the enzyme catalyzing the final step in the HBP,UAP1 (UDP-N-acetylglucosamine pyrophosphorylase 1) areupregulated in localized prostate cancer (P < 0.05; Fig. 1A).

OGT is an integration point ofHBP that is required tomodifytarget proteins via O-linked glycosylation and acts as a criticalregulator of protein stability and activity. We first confirmedthat OGT is upregulated in localized prostate cancer in themRNA level (Fig. 1A) and went on to assess the protein levelexpression of OGT in 1,987 clinically annotated prostate can-cers by IHC (Fig. 1B and Table 1). Increased OGT expressioncorrelated with increasing Gleason score and pT/pN stages(P < 0.0001), preoperative PSA (P < 0.01), and also with BCR(P < 0.0001; Table 1 and Fig. 1C).

Regulation of theHBPpathway by the androgen receptorWe further evaluated the linkage between AR and HBP gene

expression in AR-expressing prostate cancer cell lines. LNCaPcells express a mutant variant of the AR (23), whereas VCaPcells harbor the TMPRSS2-ERG fusion gene and are also

Glycosylation Regulates the Stability of c-MYC

www.aacrjournals.org Cancer Res; 73(16) August 15, 2013 5279




characterized by amplification andoverexpression ofwild-typeAR (24). LNCaP and VCaP cells were deprived of androgens for72 hours to minimize AR activity and then treated with asynthetic androgen (R1881) and mRNA was extracted after 18hours. The expression of GFPT1 and UAP1 were increased byover 2-fold in both LNCaP and VCaP cells (Fig. 2A and B). Theupregulation of GFPT1 andUAP1was confirmed at the proteinlevel in both cell lines (Fig. 2D and E).

To determine whether the hormone-dependent upregula-tion of the HBP enzymes is a feature found preferentially incancer cell lines, we used RWPE-1 cells as a contrast control.This cell line was derived from normal prostate epithelia andexpresses wild-type AR (25). RWPE-1 cells were cultured in theabsence of growth factors for 48 hours to achieve a clearresponse to androgens, stimulated with R1881 and mRNA andprotein lysates were collected. We observed no changes in theexpression of genes associated with the HBP at the transcript(Fig. 2C) or protein levels (Fig. 2F).

Targeting OGT with a small-molecule inhibitor or siRNAdecreases cell viability

Having established the cancer-specific upregulation of HBPenzymes, we wanted to inhibit this pathway and hypothesized

that the most prominent target is OGT, which is positioned tointegrate HBP activity to regulate target proteins and can betargeted with small-molecule inhibitors (18). We treatedLNCaP and VCaP cells with a concentration gradient of anOGT inhibitor (ST045849) for 48 hours and assessed theviability using an MTS assay. OGT inhibitor caused dose-dependent decrease in the viability, and the highest concen-tration (corresponding to IC50; ref. 18) caused approximately50% decrease in the viability in both cell lines (Fig. 3A and B). Inaddition, targeting OGTwith siRNA reduced the growth rate ofAR-positive prostate cancer cells (Supplementary Fig. S1A).

OGT inhibitor decreases the expression of genesassociated with DNA replication and cell cycle

The activity of OGT has been reported to be of high impor-tance for the growth of cancer cells, whereas no study hasattempted to assess the mechanism of action in an unbiasedway (6, 11–13, 26, 27). However, effects on the activity oftranscription factors through targeting OGT have previouslybeen reported (11, 26). Consequently, to understand how OGTexerts its functions in prostate cancer cells, we used expressionarrays. To minimize confounding effects from apoptosis, weselected a dose that caused maximally 20% decrease in cell

Figure 1. HBP is upregulated inhuman prostate cancer. A, mRNAsamples were collected fromprostate biopsies. In each case,both normal and cancerous tissuewas collected. The expression ofthe genes of interest wasdetermined with qPCR andnormalized to 5 housekeepinggenes (GAPDH, TBP, ALAS1,b-actin, and 18S) and then tonormal tissue. The significance ofthe expression was evaluated bypaired samples t test, �, <0.05;��, <0.01. B, evaluation of OGTstaining in localized prostatecancer. OGT staining wasclassified into 4 groups (stainingintensity 0 to þ3), which were thenused in the analysis. Examples ofstaining from each group areshown. �, positive staining isobtained from the basal andstromal cells. C, Kaplan–Meiersurvival curve plotting BCR againstpostoperative time. The probabilityof biochemical relapse–freesurvival decreased in a dose-dependent manner with increasingOGT expression (P < 0.0001,across all intensities).

Itkonen et al.





viability after 48 hours of treatment. LNCaP cells were treatedwith OGT inhibitor or siOGT for 12 hours and 24 hours tocapture the processes affected shortly after the treatment. OGT

inhibitor caused a 2-fold increase in the levels ofOGTmRNA at24 hours, whereas siOGT reduced the levels of OGT mRNA by80% (Fig 3C). Biologic triplicate samples were analyzed by

Table 1. Clinical parameters of OGT staining

Parameter n Int 0 Int 1 Int 2/3 P

Staining 1,987 425 (21.4) 1,327 (66.8) 235 (11.8)PSA<4 305 62 (20.3) 210 (68.9) 33 (10.8) 0.00974–10 1,047 232 (22.2) 710 (67.8) 105 (10)10–20 439 96 (21.9) 288 (65.6) 55 (12.5)>20 155 30 (19.4) 91 (58.7) 34 (21.9)

pT StagepT2 1,291 315 (24.4) 868 (67.2) 108 (8.4) <0.0001pT3a 401 66 (16.5) 270 (67.3) 65 (16.2)

�pT3b 277 41 (14.8) 176 (63.5) 60 (21.7)pN StageNx 1,011 239 (23.6) 699 (69.1) 73 (7.2) <0.0001N0 888 174 (19.6) 567 (63.9) 147 (16.6)N1-3 63 9 (14.3) 41 (65.1) 13 (20.6)

Gleason score�3þ3 856 235 (27.5) 563 (65.8) 58 (6.8) <0.00013 þ 4 866 167 (19.3) 587 (67.8) 112 (12.9)4 þ 3 213 18 (8.5) 149 (70) 46 (21.6)

�4þ4 34 2 (5.9) 15 (44.1) 17 (50)MarginR0 1,546 356 (23) 1,033 (66.8) 157 (10.2) <0.0001R1 422 66 (15.6) 281 (66.6) 75 (17.8)

NOTE: Samples were analyzed on the basis of the staining intensities (Int) shown in Fig. 1B. The number of patients falling into eachgroup is reported and the percentage is shown in brackets. Reported P values represent those calculated across all classes.

Figure 2. AR regulates the expression ofHBPenzymes in prostate cancer cell lines. Cell lineswere deprived of androgens for 72 hours (LNCaP andVCaP) or 48hours (RWPE-1) and stimulated with synthetic androgen (10 nmol/L R1881; ref. 47). A–C, mRNA levels after 18 hours of R1881 stimulation. Expression levelswere normalized first to TBP and then to vehicle-treated condition. Data for LNCaP and VCaP cells are obtained from 2 biologic replicates and for RWPE-1cells from2 technical replicates.D–F, protein lysateswere harvestedat the indicated timepoints after R1881stimulation andblotted for theproteinsof interest.The intensity of Western blotting signals was determined with densitometry, normalized to b-tubulin, and the amount at 0 hours was set to one.






Figure 3. OGT regulates genes associatedwith cell-cycle andDNA replication. LNCaP (A) andVCaP (B) cells were treatedwith a concentration gradient ofOGTinhibitor for 3 days and samples were analyzed with an MTS assay. The viability of vehicle-treated sample was set to 100% and other conditions werenormalized to this. The values shown are from 4 technical replicates and SEM is shown. �, P < 0.05; ��, P < 0.01; ��, P < 0.001; ��, P < 0.0001. C, LNCaPcells were treated either with siOGT or with a low dose of OGT inhibitor causing 20% decrease in the viability after 48 hours of treatment. RNA washarvested in biologic triplicates at each time point and the expression levels of genes of interest were determined with qPCR, normalized to actin and then toeither vehicle-treated condition (OGT inhibitor) or scrambled control (siOGT). D, expression profile was assayed using Illumina HT12v4 Bead Arrays.Data were quantile normalized using Bead Studio, and differentially expressed genes were identified using the JExpress software package. The 5 mostupregulated genes (DDIT3, AKR1C2, ARHGEF2, AKR1C4, and FAM129A) or downregulated genes (CDC2, c11Orf82, ATAD2, MCM10, and ASF1B) fromcells treated with OGT inhibitor for 24 hours were selected for validation of the data with qPCR. Expression of MYC mRNA was not changed by OGTinhibitor. The valueswere first normalized to TBP and then to vehicle-treated condition. E, the expression levels of the 20most downregulated genes commonto both time points after treatment with OGT inhibitor were compared with a published expression array data set (32). Within this dataset, there are atotal of 218 cases and we selected the prostatectomy samples that later on developed into metastatic disease and compared these with the cases withoutprostate cancer (PAN) based on the clinical data described in Supplementary Table S1 of Taylor and colleagues (32).

Itkonen et al.





expression arrays and data were processed using J-Express (28)to produce lists of up- and downregulated genes from eachtreatment and time point (Supplementary Table S2.). We firstevaluated the expression profiles after treating cells withsiOGT. Strikingly, MGEA5, a gene encoding for O-GlcNAcase,was the topmost downregulated gene (after OGT) at 24 hours(Supplementary Table S2 and Supplementary Fig. S1B). This isin good agreement with previous reports showing that inhi-bition ofOGT expressionwith siRNA is rapidly compensated bycommensurate loss of MGEA5 (29). MGEA5 catalyses theremoval of O-GlcNAc from OGT substrates and consequentlythis compensation effect underscores the importance of thispathway but also challenges the interpretation of OGT knock-down data.We therefore concentrated on the data generated from

cells treated with the OGT inhibitor. First, we took the 5most up- or downregulated genes from 24 hours time pointfor validation with qPCR. We observed 1.5- to 5-fold increasein the levels of DDIT3, AKR1C2, ARHGEF2, AKR1C4, andFAM129A (Fig. 3D). Next, we analyzed the expression levelsof the 5 most downregulated genes CDC2, c11orf82, ATAD2,MCM10, and ASF1B, all of which, except for c11orf82, weredownregulated by more than 50% at a 24-hour time point. Tounderstand the significance of the altered gene expressionprofile, we conducted a pathway enrichment analysis withDAVID (30, 31) gene ontology enrichment tool (Supplemen-tary Table S2). This approach revealed a consistent down-regulation of gene networks associated with cell-cycle pro-gression and DNA processing, upon treatment with OGTinhibitor.We wanted to understand the clinical relevance of OGT

inhibitor-induced changes in the expression profile, and there-fore evaluated the expression of the 20 most downregulatedgenes common to both time points in clinical samples usingdata published in Taylor and colleagues (32). Interestingly,genes downregulated by OGT inhibitor were overexpressed inthe prostate cancer tissue of patients with metastatic prostatecancer (Fig. 3E).

Inhibition of OGT destabilizes c-MYC in prostate cancercellsOGT is a known epigenetic regulator (6), and we wanted

to assess whether OGT inhibitor-induced changes in thetranscriptome are mediated by specific transcription factors.We therefore took the 200 most downregulated genes com-mon to both time points and uploaded them into theIngenuity Pathway Analysis tool (Ingenuity Systems Inc,www.ingenuity.com.) to look for common upstream regula-tors (Supplementary Table S2). The E2F family of transcrip-tion factors was the most prominent group of transcriptionfactors discovered in this analysis. We next compared theexpression levels of the candidate transcription factors withthe expression levels of OGT in a large prostate cancer geneexpression array dataset (32) and found that c-MYC is tightlycoexpressed with OGT (Supplementary Fig. S2A). We wantedto take a more stringent approach and took only the 20 mostdownregulated genes common to both time points and usedan analysis based on first neighbor associations with other

factors as reported in the REACTOME. By these criteria,c-MYC was the sole transcription factor linked to the net-work (Supplementary Fig. S2C).

Consequently, we assessed the effects of OGT inhibitor onthe stability of c-MYC. We first confirmed that OGT inhibitorwas taken up by the cells by assessing the total O-GlcNAcsignal in cell lysates after treatment with drug using blottingand densitometry. We observed a dose-dependent decreasein the levels of O-GlcNAc in 3 cell lines (LNCaP, VcaP, andPC3; Fig. 4A). The highest dose caused a minimum of 34%decrease in O-GlcNAc levels. The level of inhibition is similarto previously reported responses using the same drug (19).Next, we assessed the effects of OGT inhibitor on c-MYC andfound that inhibition of OGT results in loss of c-MYC in bothAR-positive (LNCaP and VCaP) and AR-negative (PC3) pros-tate cancer cell lines (Fig. 4A). Treatment of cells with theOGT inhibitor did not reduce c-MYC mRNA levels (Fig. 3D),suggesting that OGT regulates c-MYC stability through aposttranscriptional mechanism. OGT activity can also bereduced by targeting the enzyme with siRNA and the levelsof c-MYC decreased significantly after knocking down ofOGT for 96 hours (Supplementary Fig. S3A). c-MYC is knownto be modified by O-GlcNAc modification (33, 34) and weconfirmed this in prostate cancer cells with lectin- andimmunoprecipitation-based enrichments (SupplementaryFig. S3B and S3C).

The data from prostate cancer cell lines support a linkbetween OGT activity and c-MYC. The importance of thisoncogenic transcription factor has been shown in a range ofcancers, and amplification of the MYC locus has also beenassociated with poor prognosis in patients with prostatecancer (35). We speculated that OGT overexpression mightassociate with c-MYC copy number variation in the clinicalsetting to increase signaling via c-MYC oncogene in the lethalprostate cancer. The copy number status of MYC has beendetermined for 1,306 patients (35) in the cohort used to assessOGT expression here, which enabled us to test the potentialassociation between MYC and OGT in the clinical setting.Interestingly, we observed a statistically significant association(P ¼ 0.0012) between the increase in MYC copy number andOGT intensity (Fig. 4B). For the further validation, we went onto assess the potential coexpression between OGT andMYC inthe mRNA level. Interestingly, we found a positive correlationbetween OGT and MYC, but not between OGT and AR or OGTand ERG (Fig. 4C).

DiscussionIn this study, we used bioinformatics to identify dysre-

gulated gene networks during the early stages of prostatecancer development. This approach led to the identificationof HBP as a pathway capable of discriminating betweenbenign prostate tissue and cancerous tissue. We confirmedthe upregulation of the rate-limiting and the final enzymesin HBP in patients diagnosed with localized prostate cancer(Fig. 1A). These results are in good agreement with recentstudies reporting GFPT1 as a candidate oncogene in pan-creatic cancer (36). We also found elevated mRNA levels of






OGT in prostate cancer tissue and confirmed this result byIHC (Fig. 1B and C and Table 1). OGT and O-linked glyco-sylation have been frequently associated with cancer devel-opment (11–13, 19, 26, 37).

AR has been acknowledged as an important driver ofprostate cancer (1), and we observed that both the rate-limiting enzyme (GFPT1) and the last enzyme (UAP1) areregulated by AR (Fig. 2A–F), which enables prostate cancercells to maintain inappropriately high levels of the HBPenzymes. AR has been shown to activate multiple pathwaysin prostate cancer cells (2, 3), and the importance of ourresults lies in the identification of a confined AR-driven genemodule integrating several aspects of metabolism to regu-late cell proliferation (Fig. 5).

Targeting OGT with short hairpin RNA in mouse xenograftmodels decreases the growth of tumors (11, 13, 26). We foundthat the protein level expression of OGT correlates with highGleason score and pN/T status in prostate cancer tissue andalso with biochemical recurrence (Table 1 and Fig. 1C). Theseresults position OGT as a candidate drug target and expres-

sion arrays revealed that treatment of cells with OGT inhib-itor decreases the expression of genes associated with DNAreplication and cell-cycle progression (Supplementary TableS2). We identified c-MYC as a candidate upstream regulatorof OGT inhibitor-induced changes in gene expression, andconfirmed this association both in cell lines and in clinicalsetting (Fig. 4). The importance of c-MYC for prostate cancerhas been previously documented (38–41). Taken together,these results suggest that OGT activity integrates metabolicflux to regulate the stability of c-MYC (Fig. 5). This can beviewed as a variation into the genetic "2-hit" hypothesis (42),and we propose that increased OGT expression synergizeswith MYC copy number gain to promote prostate cancerprogression.

c-MYC is under both transcriptional and posttranslationalcontrol, and protein is known to be modified by O-linkedglycosylation (33, 34). Recent therapeutic strategies to inhibitc-MYC expression and activity in cancers have included tar-geting bromodomain-containing proteinswith small-moleculeinhibitors. One example is a drug called JQ1, developed as an

Figure 4. OGT activity is required for c-MYC stabilization. A, LNCaP, VcaP, and PC3 cells were treated with a concentration gradient of OGT inhibitor for48 hours and harvested for Western blotting. RL2 antibody was used to assess the levels of O-linked glycosylation in cell lysates. Densitometric analysis ofO-GlcNAc levels was conducted on the entire lane as previously described (48) and normalized to the intensity of actin. The value for the vehicle-treatedcondition was set to one, and the O-GlcNAc signal in each treatment condition is given under the blot as a proportion of the vehicle signal. B, evaluationof OGT staining andMYCalterations in prostatectomy samples. Sampleswere analyzed on the basis of staining intensities for OGT shown in Fig. 1B, whereasFISH ofMYC for this patient cohort has been reported previously (35). The classifications listed are "normal" (no ploidy change), "poly" (gene copy >2, ratio 1),"gain" (ratio 1-2), and amp (ratio>2). We found statistically significant coexpression between OGT expression and MYC amplification (P ¼ 0.0012). C,correlation between the mRNA level expression ofOGT and c-MYC,OGT and AR, andOGT and ERG in prostate biopsies (n¼ 17). In each case, both normaland cancerous tissues were collected. The expression of the genes of interest was normalized to 5 housekeeping genes (GAPDH, TBP, ALAS1, b-actin, and18S) and then to normal tissue.

Itkonen et al.





inhibitor of BRD4 (43). This drug reduces c-MYC expression atthe transcript level and has a significant impact on tumori-genesis in lymphoma cell lines and mouse models (44, 45). Incontrast, the OGT inhibitor we used here reduced the proteinlevel of c-MYC without impacting on c-MYC transcript expres-sion (Figs. 3D and 4A). The stabilization of c-MYC has beenreported to occur via serine/threonine phosphatase PP2A (46),and our work adds glycosylation as an alternative mechanismto stabilize c-MYC in cancer cells (Fig. 5). Future work willfocus on the development of new reagents able to discriminatebetween glycosylated and nonglycosylated pools of this impor-tant oncogene.Our data suggest that OGT activity supports the metabolic

reprogramming of tumor cells in the clinical setting by syner-gizing with MYC copy number gain to maximize c-MYCactivity. Taking further, our results support targeting OGT, orindeed c-MYC, in the treatment of prostate cancer.

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

Authors' ContributionsConception and design: H.M. Itkonen, I.G. MillsDevelopment of methodology: H.M. ItkonenAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H.M. Itkonen, S. Minner, I.J. Guldvik, M.J. Sandmann,M.C. Tsourlakis, V. Berge, T. Schlomm, I.G. MillsAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): H.M. Itkonen, S. Minner, I.J. Guldvik, V. Berge,A. Svindland, I.G. MillsWriting, review, and/or revision of the manuscript: H.M. Itkonen,I.J. Guldvik, V. Berge, A. Svindland, T. Schlomm, I.G. MillsAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): T. Schlomm, I.G. MillsStudy supervision: I.G. Mills

AcknowledgmentsThe authors thank Professor Kristin Tasken and her group for their com-

ments and critical reading of the manuscript. Microarray hybridization services

Figure 5. HBP.We used clinical gene expression data of upregulated transcripts in localized prostate cancer (20) and ChIP-seq data on AR (2) and ERG (21) toidentify a pathway that is overexpressed in the clinical setting, is targeted by AR and ERG in vitro, and integrates metabolic flux. HBP requires several keymetabolites, including glucose, glutamine, Acetyl-CoA, and UTP, which makes this pathway capable of sensing overall energy status of the cell. The fluxthrough HBP is highlighted with a gray arrow. OGT uses UDP-GlcNAc as a substrate to modify target proteins to regulate their activity, highlighted here asstabilization of c-MYC. GLUT1, glucose transporter 1; HK1/2, hexokinase-1/2; PFK1/2, phosphofructokinase 1/2; ACACA, acetyl-CoA carboxylase 1; FASN,fatty acid synthase; NUDT9, nudix (nucleoside diphosphate linked moiety X)-type motif 9; GUCY1A3, guanylate cyclase 1, soluble, alpha 3; CANT1, calciumactivated nucleotidase 1; GFTP1, glutamine–fructose-6-phosphate transaminase 1; GNPNAT, glucosamine-phosphate N-acetyltransferase 1; PGM,phosphoglucomutase; SLC35A3, solute carrier family 35 (UDP-GlcNAc transporter); G-6-P, glucose 6-phosphate; F-6-P, fructose 6-phosphate; F-1,6-BP,fructose 1,6-bisphosphate; GlcN-6-P, glucosamine-6-phosphate; GlcNAc-6-P, N-acetylglucosamine 6-phosphate; GlcNAc-1-P, N-acetylglucosamine1-phosphate.






were provided by the Helse Sor-Ost/University of Oslo Genomics Core Facility, aNational technology platform financed by the Functional Genomics Program ofthe Research Council of Norway.

Grant SupportThis work was funded by the EU FP7 Marie Curie Integrated Training

Network, PRO-NEST (H.M. Itkonen), Norwegian Research Council, Helse Sor-

Ost and the University of Oslo (I.G. Mills), and Norwegian Cancer Society (I.G.Mills).

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 February 21, 2013; revised May 8, 2013; accepted May 9, 2013;published OnlineFirst May 29, 2013.

References1. Pagliarulo V, Bracarda S, Eisenberger MA, Mottet N, Schroder FH,

Sternberg CN, et al. Contemporary role of androgen deprivationtherapy for prostate cancer. Eur Urol 2012;61:11–25.

2. MassieCE, LynchA, Ramos-MontoyaA, Boren J, Stark R, Fazli L, et al.The androgen receptor fuels prostate cancer by regulating centralmetabolism and biosynthesis. EMBO J 2011;30:2719–33.

3. Moon JS, Jin WJ, Kwak JH, Kim HJ, Yun MJ, Kim JW, et al. Androgenstimulates glycolysis for de novo lipid synthesis by increasing theactivities of hexokinase 2 and 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 2 in prostate cancer cells. Biochem J 2011;433:225–33.

4. Wellen KE, Thompson CB. A two-way street: reciprocal regulation ofmetabolism and signalling. Nat Rev Mol Cell Biol 2012;13:270–6.

5. Wellen KE, Lu C,Mancuso A, Lemons JM, RyczkoM, Dennis JW, et al.The hexosamine biosynthetic pathway couples growth factor-inducedglutamine uptake to glucose metabolism. Genes Dev 2010;24:2784–99.

6. Hanover JA, Krause MW, Love DC. Bittersweet memories: linkingmetabolism to epigenetics through O-GlcNAcylation. Nat Rev MolCell Biol 2012;13:312–21.

7. Butkinaree C, Park K, Hart GW. O-linked beta-N-acetylglucosamine(O-GlcNAc): extensive crosstalk with phosphorylation to regulatesignaling and transcription in response to nutrients and stress. Bio-chim Biophys Acta 2010;1800:96–106.

8. Capotosti F, Guernier S, Lammers F, Waridel P, Cai Y, Jin J, et al. O-GlcNAc transferase catalyzes site-specific proteolysis of HCF-1. Cell2011;144:376–88.

9. Hanover JA, Krause MW, Love DC. Bittersweet memories: linkingmetabolism to epigenetics through O-GlcNAcylation. Nat Rev MolCell Biol 2012;13:312–21.

10. Chatham JC, Marchase RB. Protein O-GlcNAcylation: a critical reg-ulator of the cellular response to stress. Curr Signal Transduct Ther2010;5:49–59.

11. Caldwell SA, Jackson SR, Shahriari KS, Lynch TP, Sethi G, Walker S,et al. Nutrient sensor O-GlcNAc transferase regulates breast cancertumorigenesis through targeting of the oncogenic transcription factorFoxM1. Oncogene 2010;29:2831–42.

12. Mi W, Gu Y, Han C, Liu H, Fan Q, Zhang X, et al. O-GlcNAcylation is anovel regulator of lung and colon cancermalignancy. BiochimBiophysActa 2011;1812:514–9.

13. Lynch TP, Ferrer CM, Jackson SR, Shahriari KS, Vosseller K, ReginatoMJ. Critical role of O-GlcNAc transferase in prostate cancer invasion,angiogenesis and metastasis. J Biol Chem 2012;287:11070–81.

14. SchlommT, Iwers L, Kirstein P, JessenB, Kollermann J,Minner S, et al.Clinical significance of p53 alterations in surgically treated prostatecancers. Mod Pathol 2008;21:1371–8.

15. Erbersdobler A, Fritz H, Schnoger S, Graefen M, Hammerer P, HulandH, et al. Tumour grade, proliferation, apoptosis, microvessel density,p53, and bcl-2 in prostate cancers: differences between tumourslocated in the transition zone and in the peripheral zone. Eur Urol2002;41:40–6.

16. Bubendorf L, Kononen J, Koivisto P, Schraml P, Moch H, Gasser TC,et al. Survey of gene amplifications during prostate cancer progressionby high-throughout fluorescence in situ hybridization on tissue micro-arrays. Cancer Res 1999;59:803–6.

17. Minner S, Jessen B, Stiedenroth L, Burandt E, Kollermann J, MirlacherM, et al. Low level HER2 overexpression is associatedwith rapid tumorcell proliferation and poor prognosis in prostate cancer. Clin CancerRes 2010;16:1553–60.

18. Gross BJ, Kraybill BC, Walker S. Discovery of O-GlcNAc transferaseinhibitors. J Am Chem Soc 2005;127:14588–9.

19. Lima VV,Giachini FR, Carneiro FS,CarvalhoMH, Fortes ZB,WebbRC,et al. O-GlcNAcylation contributes to the vascular effects of ET-1 viaactivation of the RhoA/Rho-kinase pathway. Cardiovasc Res 2011;89:614–22.

20. Gorlov IP, Byun J, Gorlova OY, Aparicio AM, Efstathiou E, LogothetisCJ. Candidate pathways and genes for prostate cancer: a meta-analysis of gene expression data. BMC Med Genomics 2009;2:48.

21. Yu J, Mani RS, Cao Q, Brenner CJ, Cao X, Wang X, et al. An integratednetwork of androgen receptor, polycomb, and TMPRSS2-ERG genefusions in prostate cancer progression. Cancer Cell 2010;17:443–54.

22. Van den Steen P, Rudd PM, Dwek RA, Opdenakker G. Concepts andprinciples of O-linked glycosylation. Crit Rev Biochem Mol Biol 1998;33:151–208.

23. Horoszewicz JS, Leong SS, Chu TM, Wajsman ZL, Friedman M,Papsidero L, et al. The LNCaP cell line–a new model for studies onhuman prostatic carcinoma. Prog Clin Biol Res 1980;37:115–32.

24. Korenchuk S, Lehr JE, MClean L, Lee YG, Whitney S, Vessella R, et al.VCaP, a cell-based model system of human prostate cancer. In vivo2001;15:163–8.

25. Bello D,WebberMM, Kleinman HK,Wartinger DD, Rhim JS. Androgenresponsive adult human prostatic epithelial cell lines immortalized byhuman papillomavirus 18. Carcinogenesis 1997;18:1215–23.

26. Lynch TP, Ferrer CM, Jackson SR, Shahriari KS, Vosseller K, ReginatoMJ. Critical role of O-Linked beta-N-acetylglucosamine transferase inprostate cancer invasion, angiogenesis, and metastasis. J Biol Chem2012;287:11070–81.

27. Ranuncolo SM,Ghosh S, Hanover JA, Hart GW, Lewis BA. Evidence ofthe involvement ofO-GlcNAc-modifiedhumanRNApolymerase II CTDin transcription in vitro and in vivo. J Biol Chem 2012;287:23549–61.

28. Hammerer P, Madersbacher S. Landmarks in hormonal therapy forprostate cancer. BJU Int 2012;110 Suppl 1:23–9.

29. Kazemi Z,ChangH,Haserodt S,McKenC, ZacharaNE.O-linked beta-N-acetylglucosamine (O-GlcNAc) regulates stress-induced heatshock protein expression in a GSK-3beta-dependent manner. J BiolChem 2010;285:39096–107.

30. Huang da W, Sherman BT, Lempicki RA. Systematic and integrativeanalysis of large gene lists using DAVID bioinformatics resources. NatProtoc 2009;4:44–57.

31. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichmenttools: paths toward the comprehensive functional analysis of largegene lists. Nucleic Acids Res 2009;37:1–13.

32. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS,et al. Integrative genomic profiling of human prostate cancer. CancerCell 2010;18:11–22.

33. Chou TY, Hart GW, Dang CV. c-Myc is glycosylated at threonine 58, aknownphosphorylation site and amutational hot spot in lymphomas. JBiol Chem 1995;270:18961–5.

34. Chou TY, Dang CV, Hart GW. Glycosylation of the c-Myc transactiva-tion domain. Proc Natl Acad Sci U S A 1995;92:4417–21.

35. El Gammal AT, Bruchmann M, Zustin J, Isbarn H, Hellwinkel OJ,Kollermann J, et al. Chromosome 8p deletions and 8q gains areassociated with tumor progression and poor prognosis in prostatecancer. Clin Cancer Res 2010;16:56–64.

36. Ying H, Kimmelman AC, Lyssiotis CA, Hua S, Chu GC, Fletcher-Sananikone E, et al. Oncogenic Kras maintains pancreatic tumorsthrough regulation of anabolic glucose metabolism. Cell 2012;149:656–70.

Itkonen et al.





37. Krzeslak A, Forma E, Bernaciak M, Romanowicz H, Brys M. Geneexpression of O-GlcNAc cycling enzymes in human breast cancers.Clin Exp Med 2012;12:61–5.

38. Hawksworth D, Ravindranath L, Chen Y, Furusato B, Sesterhenn IA,McLeod DG, et al. Overexpression of C-MYC oncogene in prostatecancer predicts biochemical recurrence. ProstateCancer ProstaticDis2010;13:311–5.

39. Gurel B, Iwata T,KohCM, JenkinsRB, LanF, VanDangC, et al. NuclearMYC protein overexpression is an early alteration in human prostatecarcinogenesis. Mod Pathol 2008;21:1156–67.

40. Sato H, Minei S, Hachiya T, Yoshida T, Takimoto Y. Fluorescence insitu hybridization analysis of c-myc amplification in stage TNM pros-tate cancer in Japanese patients. Int J Urol 2006;13:761–6.

41. Fleming WH, Hamel A, MacDonald R, Ramsey E, Pettigrew NM,Johnston B, et al. Expression of the c-myc protooncogene in humanprostatic carcinoma and benign prostatic hyperplasia. Cancer Res1986;46:1535–8.

42. Knudson AG Jr. Mutation and cancer: statistical study of retinoblas-toma. Proc Natl Acad Sci U S A 1971;68:820–3.

43. Filippakopoulos P, Qi J, Picaud S, Shen Y, Smith WB, Fedorov O,et al. Selective inhibition of BET bromodomains. Nature 2010;468:1067–73.

44. Mertz JA, Conery AR, Bryant BM, Sandy P, Balasubramanian S,Mele DA, et al. Targeting MYC dependence in cancer by inhibitingBET bromodomains. Proc Natl Acad Sci U S A 2011;108:16669–74.

45. Zuber J, Shi J, Wang E, Rappaport AR, Herrmann H, Sison EA, et al.RNAi screen identifies Brd4 as a therapeutic target in acute myeloidleukaemia. Nature 2011;478:524–8.

46. Westermarck J, HahnWC.Multiple pathways regulated by the tumorsuppressor PP2A in transformation. Trends Mol Med 2008;14:152–60.

47. Cheng J, Wang D, Wang Z, Yeh ET. SENP1 enhances androgenreceptor-dependent transcription through desumoylation of histonedeacetylase 1. Mol Cell Biol 2004;24:6021–8.

48. Ngoh GA, Hamid T, Prabhu SD, Jones SP. O-GlcNAc signalingattenuates ER stress-induced cardiomyocyte death. Am J PhysiolHeart Circ Physiol 2009;297:H1711–9.






2013;73:5277-5287. Published OnlineFirst May 29, 2013.Cancer Res Harri M. Itkonen, Sarah Minner, Ingrid J. Guldvik, et al. the Stability of c-MYC in Human Prostate Cancer CellsO-GlcNAc Transferase Integrates Metabolic Pathways to Regulate

Updated version

10.1158/0008-5472.CAN-13-0549doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/05/28/0008-5472.CAN-13-0549.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/16/5277.full#ref-list-1

This article cites 48 articles, 15 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/16/5277.full#related-urls

This article has been cited by 12 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://cancerres.aacrjournals.org/content/73/16/5277To 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-13-0549

http://cancerres.aacrjournals.org/content/suppl/2013/05/28/0008-5472.CAN-13-0549.DC1

http://cancerres.aacrjournals.org/content/73/16/5277.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/16/5277.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/16/5277


o-glcnac transferase integrates metabolic pathways to ... · o-glcnac transferase integrates...

Documents