a dna methylation prognostic signature of glioblastoma ... · brain research center, new delhi, and...

Integrated Systems and Technologies

A DNA Methylation Prognostic Signature of Glioblastoma:Identification of NPTX2-PTEN-NF-kB Nexus

Sudhanshu Shukla1, Irene Rosita Pia Patric1, Sivaarumugam Thinagararjan1, Sujaya Srinivasan1,Baisakhi Mondal1, Alangar S. Hegde2, Bangalore A. Chandramouli3, Vani Santosh4,Arimappamagan Arivazhagan3, and Kumaravel Somasundaram1

AbstractGlioblastoma (GBM) is the most common, malignant adult primary tumor with dismal patient survival, yet the

molecular determinants of patient survival are poorly characterized. Global methylation profile of GBM samples(our cohort; n ¼ 44) using high-resolution methylation microarrays was carried out. Cox regression analysisidentified a 9-gene methylation signature that predicted survival in GBM patients. A risk-score derived frommethylation signature predicted survival in univariate analysis in our and The Cancer GenomeAtlas (TCGA) cohort.Multivariate analysis identified methylation risk score as an independent survival predictor in TCGA cohort.Methylation risk score stratified the patients into low-risk and high-risk groups with significant survival difference.Network analysis revealed an activated NF-kB pathway associationwith high-risk group. NF-kB inhibition reversedglioma chemoresistance, and RNA interference studies identified interleukin-6 and intercellular adhesionmolecule-1 as key NF-kB targets in imparting chemoresistance. Promoter hypermethylation of neuronal pentraxin II(NPTX2), a risky methylated gene, was confirmed by bisulfite sequencing in GBMs. GBMs and glioma cell lines hadlow levels of NPTX2 transcripts, which could be reversed upon methylation inhibitor treatment. NPTX2 over-expression induced apoptosis, inhibited proliferation and anchorage-independent growth, and rendered gliomacells chemosensitive. Furthermore, NPTX2 repressed NF-kB activity by inhibiting AKT through a p53-PTEN-dependent pathway, thus explaining the hypermethylation and downregulation of NPTX2 in NF-kB-activated high-risk GBMs. Taken together, a 9-gene methylation signature was identified as an independent GBM prognosticatorand could be used for GBM risk stratification. Prosurvival NF-kB pathway activation characterized high-riskpatients with poor prognosis, indicating it to be a therapeutic target. Cancer Res; 73(22); 6563–73. �2013 AACR.

IntroductionThe grade IV glioma (glioblastoma; GBM) is the most

common, malignant primary brain tumor in adults. The prog-nosis remains poor with median survival ranging from 12 to 15months in spite of improvements in treatment protocol (1).Genetic heterogeneity in GBM has been proposed to explainthe limitations in the effectiveness of current therapies, whichnecessitates the need for prognostic gene signature (2). Manygenetic and epigenetic alterations aswell as expression of somegenes have been correlatedwith poor or better prognosis (3–7).In addition, molecular biomarkers like methyl guanine methyltransferase (MGMT) promoter methylation, isocitrate dehy-

drogenase 1 (IDH1) mutation status, and Glioma-CpG IslandMethylator Phenotype (G-CIMP) have been identified as GBMprognostic indicators (1, 8, 9).

Expression profiling studies have identified mRNA andmicroRNA signatures for classification and prognosis in GBM(10–14). However, none of the gene signatures have beentranslated into clinics, suggesting the need for more robustprognostic gene signature panels. Here, we have identified andvalidated a 9-gene methylation signature for GBM prognosti-cation. Multivariate analysis with all known prognostic mar-kers and signatures for GBM identified our methylation sig-nature to be an independent GBM prognostic indicator. Fur-thermore, an activated NF-kB pathway was found to beassociated with poor prognosis. We also show that neuronalpentraxin II (NPTX2), a component of methylation signaturewith risky methylation, inhibits cell growth by antagonizingNF-kBpathway through p53-PTEN activation, thus connectingthe methylation signature to NF-kB pathway.

Materials and MethodsPlasmids and reporter constructs

pCMV-Entry/NPTX2 was obtained from Origene. CAPI3K(pCDNA3-CD2p110myc) was described earlier (15, 16). CAAKT(pCDNA MYR HA AKT1) was from Addgene (Plasmid 9008).PG13-Luc was described earlier (17). CAIKK, NF-kB-Luc,

Authors' Affiliations: 1Department of Microbiology and Cell Biology,Indian Institute of Science; 2Sri SatyaSai Institute of Higher MedicalSciences; Departments of 3Neurosurgery and 4Neuropathology, NationalInstitute of Mental Health and Neuro Sciences, Bangalore, India

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

Corresponding Author: Kumaravel Somasundaram, Department ofMicrobiology and Cell Biology, Indian Institute of Science, Bangalore560012, India. Phone: 91-80-23607171; Fax: 91-80-23602697; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-13-0298

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 6563

on December 16, 2020. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 27, 2013; DOI: 10.1158/0008-5472.CAN-13-0298

http://cancerres.aacrjournals.org/

PTEN-luc, shATM, and shATR plasmid constructs wereobtained from Profs. I. Verma, K.N. Balaji, Dr. R.C.M. Simmen,Dr. Y. Shiloh, and Dr. T. De Lange, respectively.

Cell lines and reagentsTemozolomide, adriamycin, 5-aza-20-deoxycytidine, MTT,

and EscortIII transfection reagent were purchased fromSigma.Glioma cell lines U138, LN18, U343, LN229, U251, U87, T98G,U373 (all human glioma derived), 293, and C6 (Rat gliomaderived) were obtained from the laboratory of Dr. A. Guha,University of Toronto, Toronto, Canada, and grown in DMEMmedium. SVG cells were obtained from Dr. P. Seth, NationalBrain Research Center, New Delhi, and grown in MEM medi-um. The medium was supplemented with 10% FBS, penicillin,and streptomycin. No information is available about theirauthentication, although many common known mutations/alterations have been verified in our laboratory.

Patient samples and clinical dataTumor samples used for study were obtained from patients

whowere operated at Sri Sathya Sai Institute of HigherMedicalSciences and National Institute of Mental Health and Neuro-sciences, Bangalore, India. Normal brain tissue samples (ante-rior temporal lobe) obtained during surgery for intractableepilepsy were used as control samples. Tissues were bisectedand one half was snap-frozen in liquid nitrogen and stored at�80�C until DNA/RNA isolation. The other half was fixed informalin and processed for paraffin sections. These sectionswere used for the histopathologic grading of tumor andimmunohistochemical staining.

Network analysisTo find out the key signaling pathways differently regulat-

ed between low-risk and high-risk patients, we compared geneexpression profiles of The Cancer Genome Atlas (TCGA) dataset (Agilent expression data, 17,814 genes) with gene expressiondifferences between low-risk and high-risk groups by conduct-ing t-test analysis with FDR correction. We identified 3,127genes that were significantly differentially regulated betweenthe high-risk and low-risk groups (1,983 genes upregulated and1,144 genes were downregulated in high-risk compared tolow-risk). To functionally interpret the role of differentiallyregulated genes with respect to patient survival, we carried outbiological network analysis using MetaCore from ThomsonReuters, a pathway mapping tool. This tool builds biologicalnetworks from the input gene set and lists the biological proces-ses (Gene Ontology terms) associated with each network. Fromthe list of 3,127 differentially regulated genes, we chose a subsetof 61 genes, with a log2 fold change >1.5 as the input for Meta-core software. Using variant shortest path algorithm with mainparameters of relative enrichments of input genes and relativesaturation of networks with canonical pathways, Metacoreidentified 8 significant networks (Supplementary Table S3).

NF-kB gene/protein association network and landscapeanalysis of gene expression

To construct the NF-kB gene/protein interaction network,NF-kB target geneswere downloaded (18). A total of 150NF-kB

target genes were used to generate an interaction networkusing STRING database (19) using "experiment," "textmining,"and 0.700 confidence level as input option. Tomake a protein–protein interaction network, STRING curates publically avail-able databases for the information on direct and indirectfunctional protein interactions. The data file of protein–pro-tein interaction network generated from STRING was thenused in Medusa software (20) to construct a network. Thisnetwork was then analyzed by ViaComplex software (21).ViaComplex plots the gene expression activity over theMedusanetwork topology. ViaCompex overlays expression data withthe interaction network and constructs a landscape graph bydistributing microarray signal with interaction network coor-dinates. For each gene, the average of each group (high-risk andlow-risk) was plotted over the average of the same gene innormal brain and landscape was constructed.

Other methods and additional informationThe details for additional clinical data, genomic DNA extrac-

tion, sodium bisulfite conversion, methylation array analysis,bisulfite sequencing, RNA isolation, real-time quantitativeRT-PCR analysis, 5-Aza-2'-deoxycytidine treatment, transfec-tion, colony formation assays, stable cell line generation,proliferation assay, cytotoxicity assay, CHIP assay, and PTENactivity ELISA are given in the Supplementary information.

Statistical analysisTo identify the CpGs whose methylation correlates with

survival, methylation data, available for 44 patients, wereconsidered for survival analysis. The probes with missingvalue imputed using k nearest neighbor algorithm usingImputation package in R. The assessment of survival correla-tion for all 27,578 probes was performed using BRB-Array-Tools survival analysis (BRB-ArrayTools is available withoutcharge for non-commercial applications at http://linus.nci.nih.gov/BRB-ArrayTools.html). Firstly, for each CpG probe,univariate Cox proportional hazard regression was done andparametric P value was calculated. The survival time andcensoring status were randomly assigned to each patientand univariate Cox proportional hazard regression wasperformed to re-estimate the P value for each probe. Thisstep was repeated 10,000 times and permutation P value wascalculated. To select the probes with maximum variation inb value, a standard deviation criteria of >0.2 was applied.Based on these criteria, we identified 9 probes that stronglycorrelated with survival.

The methylation risk score was calculated as follows. Usingthese 9 probes, we devised a formula, resulting in amethylationrisk score that would combine their independent survivalpredicting capabilities (based on Cox regression coefficientsderived from Cox proportional hazards analysis). Each patientwas assigned a risk score that is a linear combination of b valueof the nine selected probes weighted by their respective Coxregression coefficients.

The internal validation of the risk score was carried outusing ROC analysis and subset analysis. Independent valida-tion of risk score in other data sets, multivariate analysis, andadditional details are given in the Supplementary Information.

Shukla et al.

Cancer Res; 73(22) November 15, 2013 Cancer Research6564




ResultsIdentification and validation of a 9-gene prognosticmethylation signature for GBMWith an aim to find methylation signature predictive of

GBM survival, Cox proportional hazards regression wascarried out by comparing methylation status of 27,578 CpGsand patient's (our cohort; n ¼ 44) survival data. The entireworkflow as to how the methylation signature was identi-fied, tested, and validated is shown in a schematic diagram(Fig. 1). We identified 9 probes whose methylation statuscorrelated significantly with patients' survival (Table 1 andSupplementary Materials). A methylation risk score calcu-lated for each patient by combining the effect of each of the 9genes using a risk score formula (Supplementary Materials)divided the patients into high-risk and low-risk groups withsignificant survival difference (median survival: 12 monthsvs. 23 months; Fig. 2A and Supplementary Table S1). Thestrength of methylation risk score in predicting patientsurvival was further confirmed by ROC curve analysis andsubset analysis (Supplementary Fig. S1A and SupplementaryMaterials). The methylation risk score predicted survival inGBM dataset derived from TCGA cohort as well (Fig. 2B andSupplementary Table S1). The methylation risk score alsodivided patients significantly into 3 groups in both datasetswith the identification of a third group with very high risk as

indicated by very short median survival (Supplementary Fig.S1B and S1C and Supplementary Materials).

The risk score distribution and a comparison of risk scorewith patient survival status among GBMs of our and TCGAcohort are shown (Supplementary Fig. S2A–S2D). Upon corre-lation of methylation status to patient survival, NPTX2 meth-ylation was found to be risky, whereas methylation of theremaining 8 genes was found to be protective (Fig. 2D, Table1, and Supplementary Fig. S2E). GBM tumors from high-riskpatients tend to have more methylation of NPTX2 whereastumors from low-risk group tend to have more methylation ofthe remaining genes. Additional analysis revealed that all 9genes, which form part of the methylation signature, are neededfor prognostication (Supplementary Materials).

Multivariate regression analysis indicates 9-genemethylation signature is an independent prognosticator

Cox multivariate analysis with age, we found methylation riskscore to be an independent predictor of GBM patient survival inour patient set (P ¼ 0.001; HR ¼ 1.267; B ¼ 0.237). Additionalmultivariate analysis with various prognostic factors was carriedout usingTCGAdataset. AlthoughMGMTpromotermethylation(CpG probe ID #Cg02941816), IDH1 mutation, and Glioma-CpGIsland Methylator Phenotype (G-CIMP) were found to be inde-pendent predictors when compared with age, an analysis thatincluded allmarkers identifiedmethylation risk score alone to be

GBM patients our cohort (n = 44)

(training set)

9-gene signature

Risk-score calculation

Whole genome methylation

assay

Validation-TCGA cohort

(testing set; n = 205)

Stratification into

low- and high-risk groups

Univariate and

multivariate Cox regression

analysis

Cox regression analysis and

P value cut-off

Stratification into


Univariate and


analysis

Internal validation

Area under curve

(AUC) of

risk-score

Random

subset sampling

Differential expression analysis

between low- and high-risk groups

Identification of enriched

networks and NF-kBpathway genes

Indirect validation of risk-score

Phillips et al

dataset (n = 77)

REMBRANDT

dataset (n = 157)

Unsupervised

clustering

Stratification into low-

and high-risk groupsStratification into low

and high-risk groups

Network analysis

(Metacore)

Landscape analysis for low-and high-risk compared to

normal;

Identification of NF-kBactivation in high-risk groups

STRING and

Medusa software

Landscape

analysis for low-and high-risk groups

compared to

normal group

Landscape

analysis for high-risk compared to

low-risk groups

Validation-Bent Colleagues

cohort (testing set; n = 67)

Stratification into


Univariate and


analysis

Figure 1. Schematic diagram of the entire workflow as to how the 9-gene signature was identified, tested, and validated is shown.

GBM Prognostic DNA Methylation and NPTX2-NF-kB Link

www.aacrjournals.org Cancer Res; 73(22) November 15, 2013 6565




significant with MGMT promoter methylation showing a trendtoward significance (Table 2). Interestingly, we found that allIDH1mutant patients (n¼ 9) and all G-CIMPþ patients (n¼ 11)were a subset of low-risk group (n ¼ 16; Fig. 2E).

Upon correlation with gene expression subtypes, we noticedthat although the low-risk group is highly enriched with pro-neural GBM tumors (9 of 12; 75%), the high-risk group wasmoderately enriched for mesenchymal GBM tumors (16 of 30;53%; Fig. 2E). However, survival analysis revealed that proneurallow-risk patients survived significantly longer than proneuralhigh-risk patients or all other nonproneural GBM patients (Fig.2C). On comparison with recently published 9-gene predictor(22) using TCGA dataset, we found that both signatures weresignificant in predicting survival (Supplementary Table S2). Oncomparison with 10-microRNA prognostic signature (14) usingTCGA dataset, we found the methylation risk score remainsignificant whereas the microRNA signature showed a trendtoward significance (Supplementary Table S2). Furthermore,multivariate analysis using another cohort of glioma grade III(N ¼ 67) also identified methylation risk score to be an inde-pendent survival predictor (Supplementary Table S3 and Fig. S3;ref. 23). Thus, these results put together identify methylationrisk score as an independent predictor of patient survival.

Network analysis reveals an activated NF-kB pathway inhigh-risk patients

We hypothesized that the 9 genes, which form part of themethylation signature, might modify the global gene expressionconsequently, altering some key signaling pathways differentlybetween low-risk and high-risk patients, thus explaining thedifference in their survival. Biological network analysis usingdifferentially regulated genes between low-risk and high-riskgroups derived from TCGA (Supplementary Table S4) identified8 significant networks (Supplementary Table S5). Careful anal-ysis revealed an enrichment of NF-kB pathway members inthese networks. As NF-kB pathway is associated with oncogen-esis, tumor progression, and chemoresistance in many cancers(24), we tested whether the NF-kB pathway is differentiallyactivated between low-risk and high-risk groups. First, NF-kBgene/protein interaction network was constructed using theSTRING database with 150 NF-kB target genes as input (Sup-plementary Fig. S4A). Landscape analysis of the NF-kB inter-action network using GBM patient gene expression data of low-risk versus normal and high-risk versus normal from the TCGAcohort revealed that NF-kB interaction network is very mini-mally upregulated in low-risk (as seen by very little red andorange color in z-axis plot; Fig. 3A). However, the GBM tumorsfrom the high-risk group showed much higher level of NF-kBpathway activation as evidenced by the bright red and orangecolor in z-axis plots (Fig. 3B), suggesting the possibility thathigher activation of NF-kB pathway in high-risk GBM may beresponsible for their poor treatment response and less survival.In good correlation, pretreatment with NF-kB inhibitor madeonly those cell lines, where NF-kB is already activated, morechemosensitive but not in other cell lines, wherein NF-kBactivation is very low (Fig. 3C and D and Supplementary Fig.S4B–S4D). Furthermore, silencing of p65 subunit of NF-kBrendered glioma cells more sensitive to chemotherapy (Fig. 3E).

Tab

le1.

Details

abou

t9-ge

nemethy

latio

nsign

aturethat

predicts

survival

inGBM

Our

patient

datas

etTCGAdatas

et(M

edianbva

lue)

(Med

ianbva

lue)

SerialN

o.

Sym

bol(CpG

ID)

Nam

ePva

lue

BHR

Low

risk

Highrisk

Differen

ceLo

wrisk

Highrisk

Differen

ce

1NPTX

2(cg1

2799

895)

Neu

rona

lpen

trax

in2

0.00

431.91

6.78

30.15

0.5

�0.35

0.4

0.58

�0.17

2LIMD1(cg0

4037

228)

LIM

dom

ain-co

ntaining

Proein1

0.00

18�2

.62

0.07

30.81

0.51

0.3

0.73

0.55

0.18

3MOXD1(cg1

3603

171)

DBH-likemon

ooxy

gena

seprotein

10.00

37�2

.86

0.05

70.57

0.24

0.33

0.71

0.33

0.38

4IRF6

(cg2

3283

495)

Interferon

regu

latory

factor

60.00

55�2

.23

0.10

70.5

0.29

0.21

0.56

0.41

0.16

5FB

P1(cg1

7814

481)

Fruc

tose

-1,6-bispho

s-pha

tase

10.00

56�2

.49

0.08

30.6

0.17

0.43

0.72

0.26

0.47

6FM

OD(cg2

6987

645)

Fibromod

ulin

0.00

7�2

.60.07

40.66

0.31

0.35

0.67

0.3

0.36

7LA

D1(cg2

5947

945)

Ladinin-1

0.00

81�2

.19

0.11

20.72

0.42

0.3

0.81

0.57

0.24

8RBPSUHL(cg2

1835

643)

Rec

ombiningbinding

protein

suppress

or0.00

84�2

.40.09

10.69

0.46

0.22

0.7

0.48

0.21

9GSTM

5(cg0

4987

894)

Glutathione

S-trans

ferase

Mu5

0.00

94�2

.32

0.09

80.61

0.28

0.33

0.66

0.35

0.31

Shukla et al.





It is interesting to note that several NF-kB target genes withantiapoptotic, prometastasis, prosurvival, and proinflamma-tory functions were upregulated only in GBM tumors fromhigh-risk patients (Supplementary Table S6), but were eitherdownregulated or not changed in tumors from low-riskpatients, when compared to normal, thus suggesting thepotential roles of these set of genes in imparting chemoresis-tance. Silencing each of the 7 selected target genes identifiedinterleukin (IL)-6 and intercellular adhesionmolecule (ICAM)-1 as the key NF-kB target genes to be the mediators ofchemoresistance, as their downregulation rendered gliomacells sensitive to chemotherapy consistently (Fig. 3F).Association of NF-kB with high-risk GBM was further val-

idated by means of an indirect approach using REMBRANDT(25) and Phillips (26) GBM datasets, which had both geneexpression data and patients survival information (Supple-mentary Materials). The expression levels of the same set of

genes (n¼ 3,127) that are differentially regulated between low-risk and high-risk groups in TCGA dataset (SupplementaryTable S4) was able to divide these datasets into 2 clusters withsignificant survival difference and NF-kB activation (Supple-mentary Figs. S5A and S5B and S6A–S6E). These results overallvalidate the survival prediction by our methylation signatureand confirm the association of activated NF-kB pathway withpoorer prognosis in high-risk GBMs.

NPTX2 ismethylated in GBMandhas a growth inhibitoryfunction

NPTX2 gene, a component of methylation signature, wasparticularly interesting as its methylation is risky. Bisulfitesequencing validated the hypermethylation inGBMs andgliomacell lines as against normal brain samples (Fig. 4A). NPTX2transcripts were found to be downregulated in GBM samplesand glioma-derived cell lines compared tonormal brain samples

A B

High risk (n = 32;

median survival

= 12 months)

Low risk (n = 12;

median survival

= 23 months)

P = 0.001

0 10 20 30 40 500

20

40

60

80

100

Over all survival (months) Over all survival (months) Over all survival (months)

Pe

rce

nt

su

rviv

al

0 20 40 60 80 1000

20

40

60

80

100

Pe

rce

nt

su

rviv

al

P = 0.0007

High risk (n = 168;

median survival

= 15 months)

Low risk (n = 37;

median survival

= 23 months)

0 25 50 75 1000

25

50

75

100

P = 0.028

Low risk proneural (n = 9;

median survival = 53 months) High risk proneural (n = 5;

median survival = 10 months)

Non proneural (n = 28;

median survival = 17 months)

Pe

rce

nt

su

rviv

al

C

0 0.4 0.8

D NPTX2

LIMD1

MOXD1

IRF6

FBP1

FMOD

LAD1

RBPSUHL

GSTM5

Low risk High risk

ECluster1-G-CIMP + (n = 11)

Cluster2 -G-CIMP – (n = 18)Cluster3 -G-CIMP – (n = 36)

Classical (n = 7)Proneural (n = 14)

Neural (n = 5)Mesenchymal (n = 16)

IDH1 WTIDH1 Mu

0 0.4 0.8

Color code for heat map

G-CIMP status (n = 65)

Gene expression subtypes (n = 42)

IDH1 mutation status (n = 65)

KEY

N/A (n = 23)

65 TCGA GBM samples

Glioma subtypes

IDH1 status

G-CIMP status

Low risk High risk

NPTX2

LIMD1

IRF6

FBP1FMOD

LAD1RBPSUHL

MOXD1

GSTM5

Figure 2. Risk stratification of GBM based onmethylation risk score. A and B, Kaplan–Meier graphs of (25th percentile used as cut-off) our patient cohort andTCGA cohort, respectively. C, Kaplan–Meier survival curve among proneural low-risk (green), proneural high-risk (red), and all nonproneural (blue) GBMtumors. D, heatmapofmethylation status (b value) of 9 genes frommethylation signature (our patient set; n¼44). E, heatmapofmethylation status (b value) of9 genes frommethylation signature (TCGA set; n¼ 65) for whomG-CIMP status, glioma gene expression subtypes, IDH1mutation status, andmethylation bvalues were available. D and E, the rows represent genes and columns represent patients. A color code with yellow and blue indicating high and lowmethylation, respectively, was used. The red line divides patients into low-risk and high-risk groups. E, above the heat map, each sample is additionally colorcoded as described in the KEY.






(Fig. 4B and Supplementary S7A) and methylation inhibitor,5-aza-20-deoxycytidine, treatment resulted in reexpression ofNPTX2 transcripts (Fig. 4C). NPTX2 transcripts levels inverselycorrelated with NPTX2 promoter methylation (SupplementaryFig. S7B). Further investigation revealed that exogenous over-expression of NPTX2 suppressed colony formation (Fig. 4D).NPTX2 stable clone of U343 glioma cell line, which overex-presses NPTX2 transcripts and protein (SupplementaryFig. S7C and S7D), showed increased apoptosis, reduced prolif-eration, decreased anchorage-independent growth, and increas-ed chemosensitivity compared to a vector stable (Fig. 4E–H).Furthermore, 5-aza-20-deoxycytidine pretreatment sensitizedU251glioma cells to chemotherapy in an NPTX2-dependentmanner, as simultaneous silencing of NPTX2 resulted in chemo-resistance (Fig. 4I and Supplementary S7E).

NPTX2 inhibits NF-kB pathway through p53-PTEN-PI3K-AKT–dependent manner

Next upon investigation of the connection between NPTX2and NF-kB pathway, we found NPTX2 overexpression inhibitedexpression from NF-kB luciferase reporter in glioma cell lines

(Fig. 5A), sequence-specific DNA-binding (Supplementary Fig.S8A), nuclear translocation of NF-kB subunit (p65); Supplemen-tary Fig. S9A–S9C), and repressed 6 of the 7 key NF-kB targetgenes significantly (Supplementary Fig. S8B). We also show thatNPTX2 inhibition of NF-kB could be abrogated by coexpressionof constitutively active forms of PI3 kinase, AKT, and IKKa (Fig.5B). Furthermore, NPTX2 overexpression activated PG13-Luc, ap53-dependent reporter, efficiently in U343 and U87 cells, in aconcentration-dependentmanner in 293 cells as well as in a p53-dependent manner as it failed to activate in p53 mutant U251cells (Fig. 5C) and p53 silenced cells (Supplementary Fig. S10Aand S10D). NPTX2 overexpression also induced p53 proteinlevels (but not transcript levels) and its targets, p21 and PTENprotein levels (Fig. 5E and Supplementary Fig. S8C). Aptly,NPTX2-activated expression from PTEN promoter reporterconstruct in U343 and U87 cells and in a concentration-depen-dent manner in 293 cells as well as in a p53-dependent manneras it failed to activate PTEN reporter in p53 mutant U251 cells(Fig. 5D) and p53 silenced cells (Supplementary Fig. S10B).NPTX2 also induced PTEN transcript (Supplementary Fig.S8D). More importantly, NPTX2 overexpression resulted in

Table 2. Multivariate Cox regression analysis of methylation risk score and other prognosticmarkers usingTCGA cohort

Factor No. of patients HR B (coefficient) P value

I. Univariate analysis TCGA datasetAge 205 1.033 0.033 <0.0001MGMT 205 0.226 �1.325 0.004IDH1 197 0.441 �0.916 <0.0001G-CIMP 205 0.298 �1.211 <0.0001Methylation risk score 205 1.146 0.136 <0.0001

II. Multivariate analysis with TCGA datasetAge 205 1.027 0.027 <0.0001Methylation risk score 1.109 0.103 0.005MGMT 205 0.320 �1.138 0.013Methylation risk score 1.14 0.131 <0.0001IDH1 65 0.476 �0.472 0.316Methylation risk score 1.026 0.029 <0.0001G-CIMP 205 0.400 �0.916 0.215Methylation risk score 1.030 0.030 <0.0001

III. Multivariate analysis of all the markers in TCGA datasetAge 65a NA NA 0.275MGMT NA NA 0.064IDH1 NA NA 0.322G-CIMP NA NA 0.416Methylation risk score 1.250 0.223 <0.0001

III. Multivariate analysis of all the markers in TCGA dataset (except IDH1)Age 205 1.026 0.026 0.001MGMT 0.329 �1.112 0.016G-CIMP 0.942 �0.059 0.904Methylation risk score 1.100 0.096 0.038

aMultivariate analysis that included allmarkerswas carried out in 2ways as IDH1mutationwas available only for 65patients among205patients who hadCpGmethylation values for all 9 probes that formed part of methylation signature. The analysis that includedMGMT,IDH1, G-CIMP, age, andmethylation risk score was carried out with 65 patients, whereas the analysis involving MGMT, G-CIMP, age,and methylation risk score (without IDH1) was carried out with 205 patients.

Shukla et al.





4.8-fold increase in p53 occupancy in PTEN promoter (Fig. 5F),significant activation of PTEN phosphatase activity (3–3.5 fold)as seen by increased PIP2 levels (Fig. 5G), and an efficientinhibition of AKT as seen by reduced pAKT levels (Fig. 5E).NPTX2 overexpression also decreased pIkB and increased totalIkB levels (Supplementary Fig. S9D). In good correlation, NPTX2failed to inhibit NF-kB promoter in p53 mutant U251 cells(Fig. 5A), p53 silenced cells (Supplementary Fig. S10C), and inPTEN null U87 cells (Fig. 5A). These results together suggest thatNPTX2 is a glioma growth inhibitory gene, which can modulatechemosensitivity, and its antiproliferative functions may involveinhibition of NF-kB pathway through p53-dependent PTENactivation, leading to inhibition of PI3K-AKT-IKKa signaling.

Both ATM and ATR have been shown to have overlappingand independent functions in phosphorylation and subsequentactivation of p53 during cellular genotoxic stress (27). To definethe role of ATM/ATR proteins in NPTX2 activation of p53, theability of NPTX2 to activate p53-dependent reporter activitywasmeasured in either ATMorATR silenced cells. Interestingly,we found that NPTX2 activation of p53 activity was compro-mised significantly in ATM silenced cells (Supplementary Fig.S10E and S10F) but not in ATR silenced cells (data not shown).

DiscussionIn this study, using a cohort of patients of newly diagnosed

GBM, treated uniformly and followed up prospectively, we

E

F

0

20

40

60

80

100

120

140

siControlsiLGALS3siIL6siBCL3siICAM1siIER3siCCL7siF3

U138 U343

TMZ (µmol/L)Adria (µg/mL) TMZ (µmol/L)Adria (µg/mL)N

ot

do

ne

Perc

en

t IC

50

U343 U138

Temozolomide (µmol/L)

Adriamycin (µg/mL)250

-

-

0.1

250

-

-

0.1

0

20

40

60

80

100

120

Pe

rce

nt

via

bil

ity

siControl

siNF-kB

P < 0.0001 P = 0.0009 P < 0.0001 P < 0.0001

C

BA

0

20

40

60

80

100

120

140

U343 T98G C6 U138 U251 SVG LN229

TMZ (125 µmol/L)

BAY (4 µmol/L) + TMZ (125 µmol/L)

Pe

rce

nt

via

bilit

y

NF-kB activated NF-kB not activated

P = 0 .013 P = 0.024 P = 0.0249 P < 0.001 NS NS NS

D

U343 T98G C6 U138 U251 SVG LN229

Adria (0.5 µg/mL)

BAY (4 µmol/L) + Adria (0.5 µg/mL)

Pe

rce

nt

via

bil

ity

NF-kB activated NF-kB not activated

P = 0013 P = 0.013 P = 0.01 P < .003 NS NS NS

0

20

40

60

80

100

120

140

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0X-axis X-axis

Y-ax

is

Z-axis

Z-axisY-

axis

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

Dec

reas

e In

crea

se

Dec

reas

e In

crea

se

Figure 3. NF-kB pathway association with high-risk GBM. A and B, landscape analysis of NF-kB network (Supplementary Fig. S4A) using NF-kB target geneexpression in low-risk groupcompared tonormal brain (A) andhigh-riskgroup compared tonormal brain (B).C andD, humangliomacell lineswere treatedwithBAY 11-7082 (4 mmol/L) for 12 hours and then treated with indicated drugs for 48 hours followed by viability measurement. The proportion of viable cells fromuntreated cells/Bay-treated cells was considered as 100% (NS—P > 0.05). E, U343 and U138 cells were transfected with siRNAs against NF-kB (p65)and, after 36 hours, treated with indicated drugs. The proportion of live cells was quantified after 48 hours. The proportion of viable cells from siControl wasconsidered as 100%. F, U343 and U138 cells were transfected with siRNAs against indicated genes and after 36 hours, treated with temozolomide(125, 250, 500, and1,000mmol/L) and adriamycin (0.1, 0.2, 0.4, and0.8mg/mL). After 48hours of the drug treatment, proportion of live cellswasquantified. IC50

values of control siRNA were considered as 100% and the ratio of control siRNA IC50 compared with that of indicated siRNA is shown as percent IC50.






have identified a 9-gene methylation signature using Coxproportional hazards model that can predict survival. Further-more, the methylation signature was validated in independentcohorts. Multivariate analysis using TCGA data set with allknown prognostic markers, gene signatures, and microRNAsignature identified methylation signature as an independentGBM prognostic indicator. Network analysis revealed an asso-ciation between activated NF-kB pathway and high-risk

patients, which may explain the short survival of high-riskpatients. Inhibition of NF-kB pathway made glioma cell lineswith activated NF-kB pathway alone sensitive to chemother-apy. NPTX2, a gene with risky methylation, was validated foritsmethylation and growth inhibitory functions andwas foundto inhibit NF-kB activity through its ability to induce PTEN ina p53-dependent manner. Thus, the NPTX2-p53-PTEN path-way connected the methylation signature to NF-kB pathway.

B C

Our patient set

-8

-6

-4

-2

0

2

GBMNormal

Lo

g2 r

ati

o

TCGA set

Rembrandt set

GBMNormalGBMNormal

0

1.0

2.0

3.0

4.0

5.0

U251

Lo

g2 r

ati

o

105105

64

Aza (µmol/L)

Days

U87LN18

1.5

2.5

4.5

3.5

105105

64

105105

64

No

. o

f c

ell

s (

×1

06)

U343/Vector

U343/NPTX2

0

20

40

60

80

100

U343 U251 LN229

No

. o

f c

olo

nie

s (

%)

Vector

NPTX2

0

20

40

60

80

100

120

% C

olo

nie

s

U343/vector

U343/NPTX2

U343/V

ecto

rU

343/N

PT

X2

Adriamycin

0.4 µg/mL

Temozolomide

500 µmol/L

I

U343/Vector

U343/NPTX2

P = 0.017

P = 0.013

0

50

100

150

200

250P

erc

en

t s

urv

iva

l

Control siRNA

NPTX2 siRNA

Chemo

Aza cytidine

1

+

-+

-

2

+

-+

+

3

-+

+

-

4

-+

+

+

FD

G

H

0

50

100

150

200

250

U343/Vector

U343/NPTX2

Mean

in

ten

sit

y

E

P < 0.001 P < 0.01 P < 0.001

P = 0.002 P < 0.01 P < 0.001

0.5

1.0

1.5

2.0

2.5

02 4 61Days

ANPTX2

CpG islandBSG

TSS

Normalbrain samples

GBM

Gliomacell lines

%Methylation

0.00.00.00.5

358463584260

88606249

MI: 0.13%

MI: 57.00%

MI: 64.75%

61-100%31-60%0-30%

U343/Vector

U343/NPTX2

0

20

40

60

80

100

120

Adria TemoTaxol

Pe

rce

nt

IC5

0

P = 0.0002 P = 0.0013 P = 0.005

Analysed in Infinium

Figure 4. Validation of methylation status and growth inhibitory function of NPTX2 gene. A, bisulfite sequence analysis of the NPTX2 promoter. Percentagemethylationwas established as total percentage ofmethylated cytosines from7 to 10 randomly sequenced colonies. B,NPTX2 transcript levels obtained fromRT-qPCR (this study group) or microarray data (TCGA and REMBRANDT) are plotted. C, total RNA was isolated from glioma cell lines after treatment with5aza2dC and NPTX2 transcript level was quantified by RT-qPCR. For each sample, fold change in gene expression is calculated over its mean expression inuntreated sample. D, G418-resistant colonies were selected after transfection of glioma cell lines with vector or pCMV-NPTX2. Mean colony countsare displayed as percentage of vector control. E, the active caspase-3 levels were measured using FITC-DEVD-FMK in U343/Vector and U343/NPTX2 cellsand shown as mean intensity. F, cell proliferation was measured as number of cells for U343/Vector and U343/NPTX2 clones and plotted. The difference inproliferation at different time points was found to be significant (P < 0.01). G, U343/Vector and U343/NPTX2 clones were subjected to soft agar colonyformation and the number of colonies were counted and shown keeping U343/Vector as 100%. The difference was found to be significant (P < 0.028).H, U343/Vector and U343/NPTX2 clones were treated with varying concentrations of indicated drugs (adriamycin: 0.1, 0.2, 0.4, and 0.8 mg/mL; Taxol:2, 4, 8, and 16 mmol/L; temozolomide: 125, 250, 500, and 1,000 mmol/L). At 48 hours after drug addition, the proportion of live cells was quantified by MTTassay. The proportion of viable cells from untreated cells was considered as 100%. Percent IC50 is shown. I, U251 cells were transfected with eithercontrol siRNA or NPTX2siRNA. At 48 hours posttransfection, the cells were either untreated or treated with 5aza2dC (5 mmol/L). After 24 hours of 5aza2dCaddition, the cells were treated with various amounts of adriamycin, 0.1, 0.2, 0.4, and 0.8 mg/mL or temozolomide, 125, 250, 500, and 1,000 mmol/L.After 48 hours of the drug treatment, proportion of live cells was quantified by MTT assay. The absorbance of control cells was considered as 100%. Theproportion of viable cells for indicated concentrations are shown.

Shukla et al.





The strength of our methylation signature is that multivar-iate analysis with all known prognosticmarkers and signaturesidentified it as an independent predictor of GBM survival.Although low-risk group showed enrichment of patients withIDH1 mutation, proneural gene expression subtype and G-CIMP positivity (8), methylation signature alone was identifiedas an independent survival predictor. Furthermore, compar-isons with the 9 gene (22) and 10-microRNA panels (14) alsoidentified our methylation signature as an independent sur-vival predictor. Thus, it seems that our 9-gene methylationsignature is novel, robust, specific, and the only independentpredictor of survival.To provide biological insight for survival prediction by

methylation signature, with the use of network analysis, our

study identified specific activation of NF-kB pathway in high-risk group from TCGA dataset and was further validated inREMBRANDT and Phillips datasets. Although the associationbetween NF-kB pathway activation and GBM aggressivenessand chemoresistance as well as NF-kB as a therapeutic targetis known (24, 28), our study clearly shows that NF-kB pathwayis activated particularly in high-risk group. Our data alsoshow that NF-kB inhibition could sensitize only those gliomacells having activated NF-kB pathway to chemotherapy.Furthermore, we found IL-6 and ICAM1, which are upregu-lated only in high-risk tumors, as the key mediators ofchemosensitivity. In good correlation, IL-6 and ICAM1 werefound associated with invasion, angiogenesis, tumor growth,chemoresistance, and prognosis in many cancers (29–32).

VectorA

C

D

B E

F G

NPTX2

Vector NPTX2

Vector NPTX2

Vector NPTX2

Vector NPTX2

Vector NPTX2

Vector

NPTX2

150

100

50

0

150

100

50

0

LN18 U343 T98G U251 U87

NF

-kB

Lu

c. a

cti

vit

y (

%)

Lu

c. a

cti

vit

y (

%)

NS NS NS NS NS

Vector CAPI3K CAAKT CAIKK

293

293

12 h

24 h

12 h

24 h

Ad

ria

my

cin

U343/V

ecto

r

U3

43

/NP

TX

2

Myc/

NPTX2

P53

p21

PTEN

pAKT

AKT

Actin

300

200

100

0

200

150

100

50

0

200

150

100

50

0

150

100

50

0

150

100

50

0

500

400

300

200

100

0

400

300

200

100

0

U343 U87 U251

U343 U87 U251

PG

13

Lu

c. a

cti

vit

y (

%)

PT

EN

Lu

c. acti

vit

y (

%)

293

293

NS

NS250

200

150

100

50

0

293 U343

a-IgGa-IgG a-p53 a-PTEN a-IgG a-PTEN

120

100

80

60

40

20

0

80

60

40

20

0

PIP

2 (

pm

ol)

Fo

ld e

nri

ch

men

t

Figure 5. NPTX2 inhibits NF-kB pathway through a p53-PTEN–dependent pathway. A, NPTX2 vector or control vector (2 mg) was transfected along with NF-kBluciferase reporter (1 mg) and, after 36 hours, luciferase activity was measured and plotted. B, NPTX2 vector or control vector (2 mg) was transfected along withNF-kB luciferase reporter (1 mg) and constitutively active formof either PI3K, AKT, or IKK (1 mg) and, after 36 hours, luciferase activity wasmeasured and plotted.C, NPTX2 vector or control vector (2 mg) was transfected along with pG13 luciferase reporter (1 mg) and, after 36 hours, luciferase activity was measured andplotted. For 293 cells, 0.5, 1, and 2 mg of NPTX2 vector or control vector was used. D, NPTX2 vector or control vector (2 mg) was transfected along withPTEN promoter luciferase reporter (1 mg) and, after 36 hours, luciferase activity was measured and plotted. For 293 cells, 0.5, 1, and 2 mg of NPTX2 expressionvector or control vector was used. For all experiments earlier, the luciferase activity of control vector was considered as 100%. E, cell extracts from293 cells transfected with either vector or NPTX2 overexpression construct (10 mg) or U343/vector and U343/NPTX2 stables were subjected to Western blotanalysis for NPTX2 (Myc tag), p53, p21, PTEN, pAKT, AKT, and actin proteins. F, The chromatin isolated from 293 cells transfected with vector orNPTX2 construct (20 mg) was used for immunoprecipitation with control or p53 antibody and used for PCR specific p53-binding region of PTEN. G, cell extractsfrom 293 or U343 cells transfected with vector or NPTX2 construct (20 mg) were used for immunoprecipitation using control or PTEN antibody. Theimmunoprecipitates were used to conduct PTEN activity ELISA. For all experiments, test was carried out to test the significance of the observed differencesbetween 2 conditions. �, P < 0.05; ��, P < 0.01; ��, P < 0.001.






This raises the potential use of targeting NF-kB pathway intreatment protocol specifically for those patients belongingto high-risk group.

Additional analysis of methylation signature and survivalidentified NPTX2 as a very important gene as its methylation isfound to be risky with a highermethylation in high-risk groups.NPTX2 has been shown to be methylation silenced in pancre-atic cancer, shown to induce BAX, possibly involving p53activation, and shown to downregulate cyclin D1, leading tocell-cycle arrest and apoptosis (33). Our study shows thatNPTX2 is methylation silenced in GBMs, could be reexpressedby methylation inhibitors and overexpression inhibits cellproliferation, anchorage-independent growth, and reexpres-sion sensitizes glioma cells to chemotherapy. High levels ofNPTX2 mRNA and protein have been reported in neuronalcells and in subpopulation of glial cells in normal brain (34).This may explain the reduced methylation and the consequentexpression of NPTX2 in low-risk group, which is enriched forproneural gene expression type. Our work also shows thatNPTX2 overexpression inhibits NF-kB activity, suggesting thatits functions may involve activating or inhibiting signalingpathways, which are modulators of NF-kB pathway. p53 isknown to activate PTEN, leading to inhibition of PI3K-AKTpathway (35, 36). In our study, we show that NPTX2 activatesPTEN at the level of transcription in a p53-dependent manner,leading to PI3 kinase inhibition. Furthermore, we also showthat NPTX2, through its ability to activate PTEN, inhibitsAKT, which has been shown to be a key upstream activatorof NF-kB pathway (37, 38). Although NPTX2 expressioninduces p53 protein level without a change in its transcriptlevels and may involve ATM, the exact mechanism at presentremains to be investigated. These results suggest that NPTX2inhibition of NF-kB may involve p53-PTEN activation, leadingto PI3K-AKT-IKKa inhibition. Furthermore, it explains whyNPTX2 is methylation silenced in high-risk group whereinNF-kB pathway is activated.

Thus, our study uncovers a novel pathway (throughNPTX2) of a complex interaction network between differ-ential methylation of 9 genes and alteration of global generegulation, leading to modulation of signaling, in particularNF-kB pathway, between low- risk and high-risk groups.More importantly, this work identifies the 9-gene methyla-

tion signature as an independent GBM prognostic signature.It can be used to stratify the GBM patients into low-risk andhigh-risk groups, which would facilitate individualized ther-apeutic modality. Mechanistic investigation identified acti-vated NF-kB signaling as the potential cause of poor prog-nosis in high-risk patients, suggesting NF-kB as a therapeu-tic target in particular for patients who do not respond tocurrent treatment protocols.

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

Authors' ContributionsConception and design: S. Shukla, A. Arivazhagan, K. SomasundaramDevelopment of methodology: S. Shukla, B. Mondal, A. Arivazhagan,K. SomasundaramAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A.S. Hegde, B.A. Chandramouli, A. Arivazhagan,K. SomasundaramAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): S. Shukla, I.R. Pia Patric, S. Thinagararajan,S. Srinivasan, V. Santosh, K. SomasundaramWriting, review, and/or revision of the manuscript: S. Srinivasan,V. Santosh, A. Arivazhagan, K. SomasundaramAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): I.R. Pia Patric, A.S. Hegde, B.A. Chan-dramouli, V. Santosh, A. Arivazhagan, K. SomasundaramStudy supervision: A.S. Hegde, B.A. Chandramouli, K. Somasundaram

AcknowledgmentsThe results published here are in whole or part based upon data generated

by TCGA pilot project established by the NCI and NHGRI. Information aboutTCGA and the investigators and institutions who constitute the TCGAresearch network can be found at http://cancergenome.nih.gov/. The useof datasets from Institute NC, 2005, and Phillips et al. (2006) are acknowl-edged. S. Shukla thanks CSIR for the fellowship. The authors thank ProfessorsM.R.S. Rao and P. Kondaiah for valuable suggestions. The authors thank Dr.H. Noushmehr for help in acquiring TCGA data. The authors also thank B.Thota, S. Bhargava, K. Chandrashekar, K. Prem, and B.C. Shailaja for the helprendered. K. Somasundaram is a J.C. Bose Fellow of the Department ofScience and Technology.

Grant SupportThis study was supported by a grant from DBT, government of India.The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact.

Received January 31, 2013; revised July 17, 2013; accepted August 27, 2013;published OnlineFirst September 27, 2013.

References1. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ,

Janzer RC, et al. Effects of radiotherapy with concomitantand adjuvant temozolomide versus radiotherapy alone on sur-vival in glioblastoma in a randomised phase III study: 5-yearanalysis of the EORTC-NCIC trial. Lancet Oncol 2009;10:459–66.

2. Furnari FB, Fenton T, Bachoo RM, Mukasa A, Stommel JM, Stegh A,et al. Malignant astrocytic glioma: genetics, biology, and paths totreatment. Genes Dev 2007;21:2683–710.

3. Mizoguchi M, Betensky RA, Batchelor TT, Bernay DC, Louis DN, NuttCL. Activation of STAT3, MAPK, and AKT in malignant astrocyticgliomas: correlation with EGFR status, tumor grade, and survival.J Neuropathol Exp Neurol 2006;65:1181–8.

4. Nagarajan RP, Costello JF. Epigenetic mechanisms in glioblastomamultiforme. Semin Cancer Biol 2009;19:188–97.

5. Reddy SP, Britto R, Vinnakota K, Aparna H, Sreepathi HK, Thota B,et al. Novel glioblastomamarkerswith diagnostic andprognostic valueidentified through transcriptome analysis. Clin Cancer Res 2008;14:2978–87.

6. Santosh V, Arivazhagan A, Sreekanthreddy P, Srinivasan H, Thota B,Srividya MR, et al. Grade-specific expression of insulin-like growthfactor-binding proteins-2, -3, and -5 in astrocytomas: IGFBP-3emerges as a strong predictor of survival in patients with newlydiagnosed glioblastoma. Cancer Epidemiol Biomarkers Prev 2010;19:1399–408.

7. Sreekanthreddy P, Srinivasan H, Kumar DM, Nijaguna MB, SrideviS, Vrinda M, et al. Identification of potential serum bio-markers of glioblastoma: serum osteopontin levels correlate withpoor prognosis. Cancer Epidemiol Biomarkers Prev 2010;19:1409–22.

Shukla et al.





8. Noushmehr H, Weisenberger DJ, Diefes K, Phillips HS, Pujara K,Berman BP, et al. Identification of a CpG island methylator phenotypethat defines a distinct subgroup of glioma. Cancer Cell 2010;17:510–22.

9. Parsons DW, Jones S, Zhang X, Lin JC, Leary RJ, Angenendt P, et al.An integrated genomic analysis of human glioblastoma multiforme.Science 2008;321:1807–12.

10. Freije WA, Castro-Vargas FE, Fang Z, Horvath S, Cloughesy T, LiauLM, et al. Gene expression profiling of gliomas strongly predictssurvival. Cancer Res 2004;64:6503–10.

11. Godard S, Getz G, Delorenzi M, Farmer P, Kobayashi H, Desbaillets I,et al. Classification of human astrocytic gliomas on the basis of geneexpression: a correlated group of genes with angiogenic activityemerges as a strong predictor of subtypes. Cancer Res 2003;63:6613–25.

12. Nutt CL, Mani DR, Betensky RA, Tamayo P, Cairncross JG, Ladd C,et al. Gene expression-based classification of malignant gliomascorrelates better with survival than histological classification. CancerRes 2003;63:1602–7.

13. Rao SA, Santosh V, Somasundaram K. Genome-wide expressionprofiling identifies deregulated miRNAs in malignant astrocytoma.Mod Pathol 2010;23:1404–17.

14. Srinivasan S, Patric IR, Somasundaram K. A ten-microRNAexpression signature predicts survival in glioblastoma. PLoS One2011;6:e17438.

15. Reif K NC, Thomas G, Hall A, Cantrell DA. Phosphatidylinositol 3-kinase signals activate a selective subset of Rac/Rho-dependenteffector pathways. Curr Biol 1996;6:1445–55.

16. Suvasini R SK. Essential role of PI3-kinase pathway in p53-mediatedtranscription: implications in cancer chemotherapy. Oncogene 2010;29:3605–18.

17. Das S, El-DeiryWS, SomasundaramK. Regulation of the p53 homologp73 by adenoviral oncogene E1A. J Biol Chem 2003;278:18313–20.

18. Pahl H. Activators and target genes of Rel/NF-kB transcription factors.Oncogene 1999;18:6853–66.

19. von Mering C, Jensen LJ, Snel B, Hooper SD, Krupp M, Foglierini M,et al. STRING: known and predicted protein-protein associations,integrated and transferred across organisms. Nucleic Acids Res 2005;33:D433–7.

20. Hooper SD, Bork P. Medusa: a simple tool for interaction graphanalysis. Bioinformatics 2005;21:4432–3.

21. Castro MA, Filho JL, Dalmolin RJ, Sinigaglia M, Moreira JC, MombachJC, et al. ViaComplex: software for landscape analysis of gene expres-sion networks in genomic context. Bioinformatics 2009;25:1468–9.

22. Colman H, Zhang L, Sulman EP, McDonald JM, Shooshtari NL, RiveraA, et al. Amultigene predictor of outcome inglioblastoma.NeuroOncol2009;12:49–57.

23. van den Bent MJ, Gravendeel LA, Gorlia T, Kros JM, Lapre L, Wessel-ing P, et al. A hypermethylated phenotype is a better predictor ofsurvival than MGMT methylation in anaplastic oligodendroglial braintumors: a report from EORTC study 26951. Clin Cancer Res 2011;17:1–8.

24. Nogueira L, Ruiz-Ontanon P, Vazquez-Barquero A, Moris F, Fernan-dez-Luna JL. TheNF-kBpathway: a therapeutic target in glioblastoma.Oncotarget 2011;2:646–53.

25. Institute NC. REMBRANDT home page. Available at <http://rembrandt.nci.nih.gov>. Accessed 2011 February 17, 2005.

26. Phillips HS, Kharbanda S, Chen R, Forrest WF, Soriano RH, Wu TD,et al. Molecular subclasses of high-grade glioma predict prognosis,delineate a pattern of disease progression, and resemble stages inneurogenesis. Cancer Cell 2006;9:157–73.

27. Abraham RT. Cell cycle checkpoint signaling through the ATM andATR kinases. Genes Dev 2001;15:2177–96.

28. Zanotto-Filho A, Braganhol E, Schroder R, de Souza LH, Dalmolin RJ,Pasquali MA, et al. NF-kB inhibitors induce cell death in glioblastomas.Biochem Pharmacol 2010;81:412–24.

29. Becker C, Fantini MC,Wirtz S, Nikolaev A, Lehr HA, Galle PR, et al. IL-6signaling promotes tumor growth in colorectal cancer. Cell Cycle2005;4:217–20.

30. Liu Q, Li G, Li R, Shen J, He Q, Deng L, et al. IL-6 promotion ofglioblastoma cell invasion and angiogenesis in U251 and T98G celllines. J Neurooncol 2010;100:165–76.

31. Roland CL, Harken AH, Sarr MG, Barnett CC Jr. ICAM-1 expressiondetermines malignant potential of cancer. Surgery 2007;141:705–7.

32. Schroder C, Witzel I, Muller V, Krenkel S, Wirtz RM, Janicke F, et al.Prognostic value of intercellular adhesion molecule (ICAM)-1 expres-sion in breast cancer. J Cancer Res Clin Oncol 2011;137:1193–201.

33. Zhang L, Gao J, Li L, Li Z, Du Y, Gong Y. The neuronal pentraxin II gene(NPTX2) inhibit proliferation and invasion of pancreatic cancer cells invitro. Mol Biol Rep 2010;38:4903–11.

34. Moran LB, Hickey L, Michael GJ, Derkacs M, Christian LM, KalaitzakisME, et al. Neuronal pentraxin II is highly upregulated in Parkinson'sdisease and a novel component of Lewy bodies. Acta Neuropathol2008;115:471–8.

35. Stambolic VMD, Sas D, Lin Y, SnowB, Jang Y, Benchimol S, Mak TW.Regulation of PTEN transcription by p53. Mol Cell Biol 2001;153:1369–80.

36. Song MS, Salmena L, Pandolfi PP. The functions and regulation of thePTEN tumour suppressor. Nat Rev Mol Cell Biol 2012;13:283–96.

37. Ozes ON, Mayo LD, Gustin JA, Pfeffer SR, Pfeffer LM, Donner DB. NF-kB activation by tumour necrosis factor requires the Akt serine-thre-onine kinase. Nature 1999;401:82–5.

38. Romashkova JA, Makarov SS. NF-kB is a target of AKT in anti-apoptotic PDGF signalling. Nature 1999;401:86–90.






2013;73:6563-6573. Published OnlineFirst September 27, 2013.Cancer Res Sudhanshu Shukla, Irene Rosita Pia Patric, Sivaarumugam Thinagararjan, et al.

B NexusκIdentification of NPTX2-PTEN-NF-A DNA Methylation Prognostic Signature of Glioblastoma:

Updated version

10.1158/0008-5472.CAN-13-0298doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/10/16/0008-5472.CAN-13-0298.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/22/6563.full#ref-list-1

This article cites 36 articles, 11 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/22/6563.full#related-urls

This article has been cited by 5 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/73/22/6563To 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-13-0298

http://cancerres.aacrjournals.org/content/suppl/2013/10/16/0008-5472.CAN-13-0298.DC1

http://cancerres.aacrjournals.org/content/73/22/6563.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/22/6563.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/22/6563


a dna methylation prognostic signature of glioblastoma ... · brain research center, new delhi, and...

Documents