egfr-akt-smad signaling promotes formation of glioma …preserved id3 expression during noggin...

Tumor and Stem Cell Biology

EGFR-AKT-Smad Signaling Promotes Formation of GliomaStem-like Cells and Tumor Angiogenesis by ID3-DrivenCytokine Induction

XunJin1, JinlongYin2, Sung-HakKim1, Young-WooSohn1, Samuel Beck2, YoungChangLim3,Do-HyunNam4,Yun-Jaie Choi2, and Hyunggee Kim1

AbstractAberrant activation of receptor tyrosine kinases (RTK) is causally linked to the pathobiological traits of

glioblastoma and genesis of glioma stem-like cells (GSC), but the underlyingmechanism is still unknown.Here, weshow that epidermal growth factor receptor (EGFR) signaling regulates the proliferation, angiogenesis, andacquisition of GSC characteristics by inducing inhibitor of differentiation 3 (ID3) and ID3-regulated cytokines[GRO1 and interleukins (IL)-6 and 8] induction. We found that EGFR-mediated ID3 expression was regulated bySmad5, which was directly phosphorylated by AKT. Furthermore, ID3 alone imparted GSC features to primaryastrocytes derived from Ink4a/Arf-deficient mouse, and EGFR–ID3–IL-6 signaling axis gave rise to tumor cellheterogeneity. Conversely, EGFR inhibitors suppressed EGFR-AKT-Smad5–driven induction of ID3,which led to adecrease in the tumorsphere forming ability of GSCs and U87MG cells that possess an active mutant EGFR,EGFRvIII, without obvious cytotoxic effects. However, these cells seemed to regain colonogenic ability afterremoval of the EGFR inhibitors. Together, the results delineate a novel integrativemolecularmechanism inwhichthe RTK-ID signaling pathway governs genesis and maintenance of GBM histopathologic features, such as GSCs-based tumor initiation, progression, and angiogenesis. Cancer Res; 71(22); 7125–34. �2011 AACR.

Introduction

Glioblastoma multiforme (GBM) is one of the most aggres-sive malignancies that arise from the central nervous system.Despite intensive research over several decades, the mediansurvival of patient with GBM is still less than 15 months (1).GBM is a histologically heterogeneous tumor characterized byhigh proliferation, anaplastic changes, angiogenesis, invasion,and focal necrosis (1).One of the most common signaling changes in GBM is the

aberrant activation of various receptor tyrosine kinases (RTK;ref. 2). In particular, the epidermal growth factor receptor

(EGFR) is activated in more than 50% of patients with GBMthrough a constitutively active mutation (an in-frame deletionof exons 2–7 of EGFR, referred to here as EGFRvIII) or geneamplification (3). Constitutively activated EGFR signaling con-fers increased tumorigenicity to glioma cells in vitro and in vivoby increasing proliferation and resistance to apoptosis (4);however, the precise mechanism by which this aberrant EGFRsignaling leads to the histopathologic traits of GBM is stillunknown.

Some tumors, including GBM, possess a subset of cancercells (referred to as cancer stem cells, CSC) that are capable ofasymmetrical self-renewal, tumor initiation, and multilineagedifferentiation (5). In particular, CSCs derived from GBM(referred to as glioma stem cells, GSCs) drive tumor growth,invasion, and recurrence after surgical resection, irradiation,and chemotherapy. Epidermal growth factor (EGF) and basicfibroblast growth factor (bFGF) signaling maintain the stem-ness properties of neural stem cells (NSC) andGSCs, which giverise to tumors closely resembling the histopathologic pheno-type of primary tumors (6).

Inhibitor of the differentiation (ID) family (ID1–ID4) is agroup of basic helix-loop-helix (bHLH) proteins that lack aDNA-binding domain, and they inhibit the DNA-binding activ-ity of bHLH transcription factors through heterodimerization(7). ID proteins, which are cell fate determinants, are involvedin a broad range of processes associated with tumorigenesis.For example, ID proteins promote cell proliferation by sup-pressing cell-cycle–negative regulators (8), and ID1 and ID3 arerequired for tumor angiogenesis and secondary breast tumor

Authors' Affiliations: 1School of Life Sciences and Biotechnology, KoreaUniversity; 2National Research Laboratory of Animal Cell Biotechnology,School of Agricultural Biotechnology, Seoul National University; 3Depart-ment of Otorhinolaryngology-Head and Neck Surgery, Konkuk UniversitySchool of Medicine; and 4Department of Neurosurgery, Samsung MedicalCenter and Samsung Biomedical Research Institute, Sungkyunkwan Uni-versity School of Medicine, Seoul, Republic of Korea

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

X. Jin, J. Yin, S.-H. Kim, and Y.-W. Sohn contributed equally to the work.

Corresponding Author: Hyunggee Kim, School of Life Sciences andBiotechnology,KoreaUniversity,Seoul 136-713,RepublicofKorea.Phone:82-2-3290-3059; Fax: 82-2-953-0737; E-mail: [email protected] Xun Jin. Phone: 82-3290-3472; Fax: 82-2-953-0737; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-11-1330

�2011 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 7125

Research. on February 19, 2020. © 2011 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 5, 2011; DOI: 10.1158/0008-5472.CAN-11-1330

http://cancerres.aacrjournals.org/

reinitiation in lung metastasis (9, 10). We and others haverecently shown that ID4 induces brain CSC genesis throughdedifferentiation of Ink4a/Arf-deficient mouse astrocytes (11)and that ID1 and ID3 induced by TGF-b signaling mediate thetumor-initiating capacity of GSCs (12).

These findings suggest that there is a crucial link betweenRTKs and the ID family in glioma biology; however, little isknown about the signaling responsible for GSC genesis andmaintenance. In this study, we show that EGFR signaling leadsto acquisition of GSC characteristics and angiogenesis byinduction of ID3 and ID3-regulated cytokines through AKT-dependent activation of Smad5.

Materials and Methods

Cell lines, cell growth, neurosphere formation, andcolony formation on semisolid medium

The human glioma cell lines (U87MG, A1207, and U373MG)and primary human umbilical vein endothelial cells (HUVEC)were purchased from American Type Culture Collection(ATCC) and were maintained for fewer than 6 months afterreceipt. These cell lines were authenticated by ATCC andthrough various tests. Mouse astrocytes were isolated fromthe cerebral cortices of 5-day-old Ink4a/Arf-knockout mice asdescribed previously (13). X01, X02, and X03 GSC lines are GSClines established from human brain tumors (14). Cell growthand neurosphere formation assay was conducted as describedpreviously (11). Colony formation on semisolid medium wascarried out as described early (15).

Subcutaneous and orthotopic implantation assayTo establish subcutaneous xenograft models, cells (2� 106)

were subcutaneously injected into nude mice (BALB/c nu/numice). For the orthotopic implantation, 5 � 104 cells werestereotactically injected into the brain of nude mice (coodi-nates: anterior-posterior, þ2.5; medial-lateral, þ2; dorsal-ven-tral,–3 mm from the bregma). All mouse experiments wereapproved by the animal care committee at the College of LifeScience and Biotechnology, Korea University and were carriedout in accordance with government and institutional guide-lines and regulations.

StatisticsFor the multidata set experiments, ANOVA tests were con-

ducted with the Statistical Package for the Social Sciencessoftware (version 12.0; SPSS). a,b,c,dP < 0.05 or A,B,C,DP < 0.01 wasconsidered to represent a statistically significant difference.For the 2 data set experiments, 2-tailed Student t test wascarried out. �, P< 0.05 or ��, P< 0.01was considered statisticallysignificant.

Results

EGFR-AKT-Smad1,5,8 signaling axis induces ID3 inhuman glioma cells and GSCs

Aberrant activation of EGFR signaling is causally linked togliomagenesis (2), and constitutively activated EGFR signalingimparts stem cell-like characteristics in Ink4a/Arf�/� mouse

astrocytes (13). However, the underlying mechanism of EGFR-driven acquisition of GSC features is still unknown. GSCs sharea number of traits with normal NSCs, which are governed bystem cell fate determinants. In this study, we focused on IDfamily (ID1–ID4) to evaluate their roles in EGFR-driven acqui-sition of GSC properties. We first determined the expression of4 ID genes inmany types of human brainmalignancies with themining Repository of Molecular Brain Neoplasia Data (REM-BRANT) database (http://caintegrator.nci.nih.gov/rembrant).Mean expression levels of ID3 were more than 7-fold higher inastrocytomas (n¼ 148), GBMs (n¼ 226), and mixed tumors (n¼ 11) than normal brain tissue (n ¼ 28; Supplementary Fig.S1A). We also found that ID3mRNA levels were increased in 12of 15 GBM specimens (Supplementary Fig. S1B). These datasuggest that ID3 is themost relevant of the 4 ID genes in gliomapathogenesis. We then compared the expression of 4 ID geneswith EGFR expression in human gliomas with a web-basedbioinformatics database obtained from The Cancer GenomeAtlas database of National Cancer Institute (TCGA). Of the 4 IDgenes, ID3mRNA expression showed the strongest associationwith EGFR genomic amplification (80%; ID3 mRNA increasedin 53 of 66 EGFR-amplified glioma specimens) and overexpres-sion (98%; in 164 of 168 EGFR-upregulated glioma specimens;Supplementary Fig. S2A). To determine whether EGF signalingis directly associatedwith ID3 expression, 3 glioma cells, A1207,U373MG, and U87MG, were treated with EGF. EGF treatmentconsistently increased ID3 expression in 3 glioma cells withincreases in AKT and extracellular signal—regulated kinase(ERK) phosphorylation (Supplementary Fig. S2B). We alsomeasured ID3 expression in U87MG cells transduced with 2different types of constitutively activated EGFR mutants(EGFRvIII, common in human glioma; and EGFRL858R, com-mon in human lung cancer) and wild-type EGFR and foundthat ID3 protein levels were elevated by both mutant and wild-type EGFR genes (Fig. 1A). In a previous study, ID1 and ID3transcription was shown to be upregulated by phosphorylatedSmad1,5,8 (p-Smad1,5,8; ref. 16); therefore, we used a luciferasereporter assay with the ID3 promoter construct (containing aSmad-binding element). EGF promotes transcription of theID3 promoter in a dose-dependent manner, but this effect wascompletely suppressed by dorsomorphin, a chemical inhibitorof p-Smad1,5,8 (Supplementary Fig. S2C; ref. 17).We also foundthat ectopic expression of EGFRvIII in U87MG cells increasedp-Smad1,5,8, and conversely, dorsomorphin completely dimin-ished ID3 and p-Smad1,5,8 protein and mRNA expression (Fig.1B, top; and Supplementary Fig. S2D), suggesting that EGFRvIIIregulates ID3 expression through p-Smad1,5,8. However, pre-vious studies have shown that BMP induces ID expressionthrough BMP receptor (BMPR)–mediated phosphorylation ofSmad1,5,8 (16) and induces differentiation of GSCs to non-tumorigenic differentiated glioma cells (18). To determinewhether EGF signaling requires BMP/BMPR signaling toinduce ID3 and p-Smad1,5,8, we treated cells with the BMPantagonist noggin (19), which strongly reduced p-Smad1,5,8levels in U87MG-Puro control cells, but did not completelydeplete p-Smad1,5,8 levels in U87MG-EGFRvIII cells (Fig. 1B,bottom). Furthermore, although ID3 expression was weaklyreduced by noggin, EGFRvIII overexpression effectively

Jin et al.

Cancer Res; 71(22) November 15, 2011 Cancer Research7126




preserved ID3 expression during noggin treatment (Fig. 1B,bottom), indicating that EGFRvIII might regulate ID3 expres-sion and Smad1,5,8 phosphorylation in a BMP/BMPR signal-ing-independent manner. To evaluate the plausible mecha-nism underlying EGFRvIII-driven induction of ID3 and p-Smad1,5,8, AKT and ERK signaling in EGFRvIII-overexpressingU87MG cells was inhibited with LY294002 and U0126, respec-tively. Inhibition of AKT, but not ERK, reduced p-Smad1,5,8and ID3 levels (Fig. 1C), and conversely, ectopic expression ofmyristoylated AKT (myr-AKT; a constitutively active AKT) inU87MG cells increased p-Smad1,5,8 and ID3 levels (Fig. 1D).These results indicated that EGFR-AKT signaling regulates ID3expression throughp-Smad1,5,8.

AKT phosphorylates Smad5 and subsequently inducesID3 expressionTo determine how EGFR-AKT signaling induces p-

Smad1,5,8, a co-immunoprecipitation assay was conductedand Flag-tagged myr-AKT was shown to specifically bind toendogenous Smad1,5,8 (Fig. 2A, left). A previous study showedthat Smad1 and Smad5 played distinct and even opposite rolesin embryogenesis (20); thus, the expression levels of eachSmad1,5,8 was measured by real-time reverse transcriptasePCR (RT-PCR). In this analysis, Smad5 was found to be themost abundant Smad in U87MG-EGFRvIII cells (Supplemen-tary Fig. S3A). Each Smad1,5,8 was depleted using an siRNA-

mediated knockdownmethod. Although all 3 Smad1,5,8mRNAlevels decreased by more than 50% in U87MG-EGFRvIII cellstransduced with each Smad1,5,8-specific siRNA, both total andp-Smad1,5,8 and ID3 protein levels only decreased in the cellstransduced with Smad5-specific siRNA (Fig. 2A, middle; Sup-plementary Fig. S3B), suggesting that Smad5 might be a majorregulator of ID3 in EGFR-AKT signaling axis inU87MGcells. Toevaluate whether AKT directly induces Smad5 phosphoryla-tion, an in vitro kinase assay was conducted and an active myr-AKT, but not a dominant-negative AKT mutant (K179M), wasshown to enable phosphorylation of glutathione S—transfer-ase (GST)-Smad5 (Fig. 2A, right). We also found that AKT-mediated phosphorylation of Smad5was specific due to failurein phosphorylation of GST-Smad1 by myr-AKT (Fig. 2A, right).Collectively, these findings indicate that EGFRvIII-AKT signal-ing regulates Smad5 phosphorylation and subsequent ID3expression.

To further characterize the EGFR-ID3 signaling pathway inGSC biology, 3 differentGSCs, X01, X02 andX03 (14), were used.These GSCs exhibit higher levels of ID3 and p-AKTbecause X01and X03 express the mutant EGFRvIII and X02 is PTENdeficient (Fig. 2B, top). Consistent with results obtained withglioma cells, LY294002 reduced levels of p-AKT, p-Smad1,5,8,and ID3 in X01 and X02 cells (Fig. 2B, bottom). The shorthairpin RNA (shRNA)-mediated knockdown of ID3 in theseGSCs showed a significant decrease in the neurosphere form-ing ability, which is a hallmark of GSCs (Fig. 2C). Furthermore,

U87MGA

B C DU87MG

U87MG

U87MG

Noggin

EGFR

EGFR

8

6

4

2

0EGFR

U87MG-Puro

U87MG-EGFRvIII

U87MG-EGFRvIIIDorsomorphin

Mock LY UCtr.

Myc-Myr-Akt1

Myc

U87MG-EGFRL858R

U87MG-wtEGFR

Ban

d in

tens

ity

p-ERK

p-ERK

ERK

p-ERK

p-AKT

p-AKT

p-AKT

AKT

AKT

p-AKT

ID3

ID3

p-Smad1,5,8

Smad1,5,8

ID3α-Tubulin

β-Actin

EGFR

ID3

p-Smad1,5,8

Smad1,5,8

β-Actin

EGFR

p-Smad1,5,8

Smad1,5,8

ID3

β-Actin

p-Smad1,5,8

p-GSK-3β

Smad1,5,8

ID3

β-Actin

Puro Puro PuroEGFRvIII EGFRL858R

Puro PuroEGFRvIII EGFRvIII

Puro PuroEGFRvIII EGFRvIII

wtEGFR

Figure 1. EGFR signaling induces expression of ID3. A, ID3 protein level was elevated in U87MG cells transduced with constitutively activated EGFRmutants(EGFRvIIIorEGFRL858R) andwild-typeEGFR. Theprotein levelswerequantifiedby ImageJ software (right). B, top, EGFRvIII increased ID3protein expressionin U87MG cells by activating Smad1,5,8 (C-terminal phosphorylation). p-Smad1,5,8 in U87MG-Control and U87MG-EGFRvIII cells was inhibited bydorsomorphin (5 mmol/L, 4 h). Bottom, EGFRvIII-driven ID3 induction was not suppressed by noggin (50 ng/mL, 4 h). C, AKT activity in U87MG-EGFRvIII cellswas specifically required for p-Smad1,5,8–mediated ID3 expression, as assessed by treatment with the MEK/ERK inhibitor, U0126 (U: 20 mmol/L), and thePI3K/AKT kinase inhibitor, LY294002 (LY: 20 mmol/L). D, ectopic expression of myr-AKT in U87MG cells activated Smad1,5,8 and induced ID3 expression.MEK, mitogen-activated protein/ERK; Ctr, control.

ID3-Mediated Glioma Stem-like Cell Genesis and Angiogenesis

www.aacrjournals.org Cancer Res; 71(22) November 15, 2011 7127




inhibition of p-Smad1,5,8, and the concomitant decrease of ID3in X01 and X03 cells by dorsomorphin completely suppressedneurosphere formation (Fig. 2D). These findings indicate thatID3 and p-Smad1,5,8 play a pivotal role in GSC maintenance.

ID3 imparts GSC-like features to Ink4a/Arf�/� mouseastrocytes

To directly address whether ID3 alone is sufficient to giverise to GSCs, ID3 was introduced into primary Ink4a/Arf�/�

mouse astrocytes by retroviral transduction. ID3 overexpres-sion accelerated cell proliferation with marked increase incyclin E protein levels, which led to cell aggregation in culturemedium supplemented with 10% and 0.5% FBS (Fig. 3A), andproduced significant increases in in vitro colony formation on asemisolid medium and in vivo tumor formation by subcuta-neous implantation (Fig. 3B). Our previous study showed thatectopic expression of ID4 in Ink4a/Arf�/�mouse astrocytes ledto GSC genesis (11); thus, we investigated the potential role ofID3 in neurosphere formation and stemcellmarker expression.When cultured in serum-freemedium supplemented with EGFand bFGF, ID3-overexpressing cells formed neurospheresexpressing a number of GSC markers CD133, Nestin, and Sox2but not the differentiated cell lineage markers S100b (astro-cyte), Tuj1 (neuron), and O4 (oligodendrocyte; ref. Fig. 3C).When cultured in the presence of serum, ID3-overexpressingcells grown in adherent culture showed increased Nestin and

Sox2 levels and decreased glial fibrillary acidic protein levels(Supplementary Fig. S4), indicating ID3 overexpressioninduced dedifferentiation of Ink4a/Arf�/� mouse astrocytes.

Given that GSCs are causally linked to GBM heterogeneity,we assessed whether tumors arising from Ink4a/Arf�/� astro-cytes-ID3 cells retain their histopathologic features by immu-nostaining tumor sections with antibodies against NSC anddifferentiated cellmarkers. In sharp contrast to control tumorsconsisting primarily of S100bþ astrocytes, ID3-driven tumorsretained CD133þ, Nestinþ, S100bþ, and O4þ cells, as well asmicro- andmacrovasculature structures surrounded byCD31þ

endothelial cells (Fig. 3D). These results indicate that ectopicexpression of ID3 is sufficient to give rise to histologicallyheterogeneous gliomas with angiogenesis.

ID3 regulates angiogenesis and GSC genesis by cytokineinduction

Glioma angiogenesis provides a perivasculature niche thatallows GSCs to maintain their stemness (21, 22). To determinea possible role of ID3 in angiogenesis (23) and tumor cellheterogeneity, cytokine/chemokine secretion was measuredby a custom angiogenesis cytokine antibody array in U87MG-ID3 and control cells, and 3 cytokines were shown to beupregulated (both expression and secretion) by ID3 overex-pression; growth-regulated alpha protein (GRO1, also knownas CXCL1), interleukin 6 (IL-6), and interleukin 8 (IL-8)

EGFR

Flag

Flag

293T SmadCtr. Scr. siRMA1 5 8Flag-Myr-Akt1

Flag

-Myr

-Akt

1

Ctr.

Flag

-DN-

Akt1(

K179

M)

GSCs

PTEN

ID3

ID3

Dor.

P32-Smad5

P32-Smad1

WB: Flag

Dor.+–+–

+–+–

ID3

ID3

β-Actin

β-Actin

β-ActinID3

β-Actin

β-Actin

NHA XO1

XO1

M LY M LY

XO2

XO2

XO1

30

25

20

15

10

5

0

Neu

rosp

here

s

Neu

rosp

here

s

35

30

25

20

15

10

5

0

XO2

XO1 XO2 XO1 XO3

XO1 XO3

XO3

p-Smad1,5,8

AKT

p-AKT

Smad1,5,8

Smad1,5,8

Smad1,5,8 IP: Flag

Input

Smad1,5,8

p-Smad1,5,8

p-Smad1,5,8

20.0 ± 2.3

26.7 ± 2.7

0 ± 0 0 ± 0

shScr.

shID3

shScr. shID3 shScr. shID3

A

B C D

Figure 2. AKT directly regulates Smad1,5,8 activity that involves GSC maintenance. A, left, immunoprecipitation assay revealed that Flag-tagged myr-AKTinteractswith Smad1,5,8.Middle, an siRNA-mediated knockdownof Smad5 decreased ID3 expression. Right, in vitro kinase assay revealed that Flag-taggedmyr-AKT specifically phosphorylates recombinant GST-Smad5. B, top, levels of EGFRvIII, PTEN, and ID3 in 3 GSC lines derived from patients with glioma.NHA, human astrocytes. Bottom, inhibition of AKT activity in X01 and X02 by LY294002 (LY) decreased p-Smad1,5,8 and ID3 levels. C, the shRNA-mediatedknockdown of ID3 decreases neurosphere formation in X01 and X02 cells (200 cells/12-well plate). Data are expressed asmean�SD. �,P < 0.05; ��,P < 0.01.D, inhibition of p-Smad1,5,8 in X01 and X03 cells by dorsomorphin (Dor) suppressed ID3 expression and neurosphere formation. Data are expressed asmean� SD. ��, P < 0.01. Ctr, control; WB, Western blot; IP, immunoprecipitation.

Jin et al.





(Fig. 4A). To evaluate their angiogenic effects, HUVECs wereexposed to conditioned medium (CM) from U87MG cellsoverexpressing ID3 (U87MG-ID3), GRO1, IL-6, and IL-8 orU87MG-ID3 cells deficient in GRO1, IL-6, and IL-8 by ashRNA-mediated knockdown. We found that HUVEC in vitrovessel formation wasmarkedly increased after exposure to CMfrom U87MG cells overexpressing GRO1, IL-6, IL-8, and ID3(Fig. 4B). Conversely, depletion of GRO1, IL-6, and IL-8 in theU87MG-ID3 cells suppressed vessel formation (Fig. 4B), indi-cating that ID3-driven angiogenesis in vitro and in vivo (Fig. 3D)is regulated by angiogenic cytokines. In addition, it is worthnoting that although EGF is a weaker angiogenic factor than

bFGF when added directly to HUVECs; however, the CMderived from U87MG cells treated with EGF led to higher invitro tube formation of HUVECs than direct treatment of EGFto HUVECs, presumably due to EGFR-ID3 signaling-inducedcytokines (Supplementary Fig. S5A). To address this possibility,in vitro tube formation of HUVECs were compared using CMfrom U87MG-EGFRvIII cells treated with gefitinib (EGFRinhibitor), LY294002 [phosphoinositide 3-kinase (PI3K)/AKTinhibitor], and dorsomorphin (Smad1,5,8 inhibitor). In con-trast to CM from U87MG-EGFRvIII cells, in vitro tube forma-tion of HUVEC was dramatically decreased by CM fromU87MG-EGFRvIII cells treated with EGFR, AKT, and Smad

A B

C

D

200

160

120

80

40

0

35

30

25

20

15

10

5

0

1,400

1,200

1,000

800

600

400

200

00 9 12 15 18 21 24 27 30 33

Days after injection

Foc

i num

ber

Tum

or v

olum

e (m

m3 )

Puro

Puro

Puro

ID3

ID3

Puro ID3

Puro

Puro

ID3

ID3

Pur

oID

3

CD31+CD133+ID3+ Nestin+ O4+S100β+

ID3

Puro

706050403020100

Neu

rosp

here

s

ID3

0 1 2 3 4 5

Days in culture

ID3pRb

Cdk4Cdk2Cyclin D1Cyclin E

Cyclin Aα-Tubulin

10% 0.5%FBS FBSAst Ink4a/Arf –/–

AstInk4a/Arf –/–


Ast Ink4a/Arf –/–


Ast

rocy

tes

Ink4

a/A

rf–/

–C

ell n

umbe

r (×

104

)

Figure 3. ID3 switches Ink4a/Arf-deficient mouse astrocytes to GSCs. A, left, cell proliferation of AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells grown inDulbecco's Modified Eagle's Medium (DMEM) þ 10% FBS. Data are expressed as mean � SD. Inset, representative images of cell grown to confluence inDMEM supplemented with 10% FBS or 0.5% FBS. Scale bar, 20 mm. Right, protein expression levels of cell-cycle regulators in AstInk4a/Arf�/�-control andAstInk4a/Arf�/�-ID3 cells. B, top, number of transformed foci of AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells grown on semisolid medium. Data areexpressed as mean � SD. ��, P < 0.01. Right, tumorigenic potential of AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells was determined by in vivosubcutaneous transplantation (n¼6). Tumor volumesare expressed asmean�SD. Inset, representative imagesof tumors derived fromAstInk4a/Arf�/�-controland AstInk4a/Arf�/�-ID3 cells. �,P < 0.05. C, left, number of neurospheres (>10 mm) in AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells (400 cells/12-well plate)grown for 14days in serum-freemedium supplementedwith EGF andbFGF. ��,P< 0.01. Immunofluorescence images (magnification, 100�; scale bar, 50 mm)showing various stem cell and differentiated cell markers in cryomicrodissected neurospheres derived from AstInk4a/Arf�/�-ID3 cells (right). Right, tumorigenicpotential of AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells was determined by in vivo subcutaneous transplantation (n¼ 6). Tumor volumes, mean � SD.Representative photos (inset; scale bar, 2 mm) showing tumors derived from AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells. D, immunohistochemicaldetection of ID3þ, CD133þ, Nestinþ, S100bþ, O4þ, and CD31þ cells in tumors derived from AstInk4a/Arf�/�-control and AstInk4a/Arf�/�-ID3 cells. Scale bar,50 mm.






inhibitors (Supplementary Fig. S5B). These results suggest thatEGFmight play amore important role in angiogenesis throughEGFR-ID3 signaling-mediated induction of angiogenic cyto-kines in tumor cells.

Wenext assessed the neurosphere-forming ability of U87MGcells to determine the biological effect of these genes in GSCgenesis. Overexpression of ID3, IL-6, IL-8, and EGFRvIII(U87MG-EGFRvIII), but not GRO1, promoted neurosphere

formation, and depletion of IL-6 and IL-8 in U87MG-ID3 cellsor knockdown of ID3 in U87MG-EGFRvIII cells inhibitedneurosphere formation (Fig. 4C). Similar to previous studies,which reported that EGFR-driven overexpression of IL-6 pro-motes breast CSC formation through STAT3-mediated activa-tion of Jagged-Notch signaling (24), we found thatmRNA levelsof genes related to Jagged-Notch signaling (Jagged1, Hes1, andHey1) and NSC markers (CD133 and Nestin) significantlyincreased with ID3 and IL-6 overexpression. However, IL-8overexpression only weakly induced Hey1, CD133, and Nestinexpression, and GRO1 overexpression increased only Hey1mRNA levels (Fig. 5A). We also found that shRNA-mediatedknockdown of ID3 in the X01 and X02 GSC lines decreasedCD133,Nestin, IL6, IL8, Jagged1,Hes1, andHey1 expression (Fig.5B). These findings provide evidence that EGFRvIII and EGFR-driven ID3 induction in glioma cells confers GSC features, atleast in part, through IL-6–Jagged-Notch signaling axis.

EGFR–ID3–IL-6 signaling axis regulates tumor cellheterogeneity

We then focused on the biological function of the EGFR–ID3–IL-6 signaling pathway in GSC-driven tumor progressionand heterogeneity. An ELISA was carried out to measure IL-6secretion in U87MG cells treated with EGF and bFGF and inU87MG cells overexpressing ID3, H-Ras, EGFRL858R, andEGFRvIII. In agreement with previous reports (25, 26),EGFRL858R, H-Ras, and bFGF increased IL-6 secretion, as didEGF treatment and overexpression of ID3 and EGFRvIII (Sup-plementary Fig. S6A). To further assess whether IL-6 expres-sion and secretion were dependent on ID3 upstream signalingpathway, U87MG-EGFRvIII cells were treated with gefitinib,LY294002, and dorsomorphin to inhibit EGFR-AKT-Smad sig-naling pathway. We found that these inhibitors significantlysuppressed IL-6 expression and secretion (Supplementary Fig.S6A and S6B). A previous study showed that activated STAT3increased IL-6 expression in an autocrine manner. Therefore,IL-6 expression in U87MG-EGFRvIII cells was examined afterinhibiting ID3 and Janus—activated kinase (JAK)/STAT (27).We found that shRNA-mediated ID3 depletion, JAK inhibitor(P6), and STAT3 inhibitor (STAT3-I, an inhibitory peptideagainst STAT3 dimerization) decreased IL-6 mRNA expressionby 96%, 24%, and 16%, respectively (Supplementary Fig S6C),indicating that ID3 is a major regulator of EGFR-driven IL-6expression. However, because ID3 is a transcriptional repres-sor (7), these observations also raise the question of how doesID3 promote IL-6 expression. We found that ID3 overexpres-sion inhibits transcription of the IL-6 transcriptional repressorBCL6 (28), which contains an ID3-repressive element (E-box) inits own promoter, as evidenced by suppression of BCL6-promoter luciferase activity (Supplementary Fig. S6D).

To investigate the effect of the EGFR–ID3–IL-6 signaling axison tumorigenicity and heterogeneity, we injected the followingcells into immunocompromised mice: U87MG-IL-6, U87MG-EGFRvIII, U87MG-ID3, ID3-depleted U87MG-EGFRvIII, and IL-6–depleted U87MG-ID3. EGFRvIII, ID3, and IL-6 acceleratedtumor formation, whereas suppression of ID3 and IL-6 in theU87MG-EGFRvIII and U87MG-ID3, respectively, significantlyreduced tumor formation (Fig. 6A). As shown in Fig. 6B, cells

180

150

120

90

60

30

0

45

40

35

30

25

20

15

10

5

0

28

24

20

16

12

8

4

0

U87MG cells

Puro-shScr.

Gro1-shScr.

IL6-shScr.

IL8-shScr.

ID3-shScr.

ID3-shGro1

ID3-shIL8

bCDEbCDE

BCDEBCDE

ACDEFh

ABFGH

ABDEFGHIJ

CDEGHCDEGH

ABCFIJ ABCFIJ

CDEGH CDEGHCDEGH

ABCFIJABCFIJ

ABFGH

ABFGH

ID3-shIL6

Puro-shScr.

Gro1-shScr.

IL6-shScr.

IL8-shScr.

EGFRvIIIshScr.

EGFRvIII-shID3

ID3-shScr.

ID3-shGro1

ID3-shIL8

ID3-shIL6

U87MG cells

Tube

num

ber

Neu

rosp

here

s

U87MG-PuroU87MG-ID3

Gro1

1, Gro1; 2, IL6; 3, IL8

2

1

3

A

B

C

IL6 IL8

U87

MG

-Pur

oU

87M

G-I

d3 mR

NA

exp

ress

ion

Figure 4. EGFR-ID3 signaling induces cytokines that promote endothelialtube and neurosphere formation. A, left, U87MG-ID3 cells showincreased GRO1, IL-6, and IL-8 secretion as determined by theangiogenesis antibody array. Right, high levels of GRO1, IL-6, and IL-8mRNA in theU87MG-ID3 cellswere confirmedby real-timePCRanalysis.Data are expressed as mean� SD. ��, P < 0.01. B, in vitro tube formationwas elevated in HUVECs exposed to CM from U87MG-GRO1, U87MG-IL-6, U87MG-IL-8, and U87MG-ID3. The knockdown of GRO1, IL-6, andIL-8 in U87MG-ID3 cells suppressed ID3-mediated induction of tubeformation. a,b,c,d,e,f,g,h, P < 0.05; A,B,C,D,E,F,G,H, P < 0.01. C, neurosphereformation assay. EGFRvIII, ID3, IL-6, and IL-8 increased neurosphereformation of U87MG cells. A,B,C,D,E,F,G,H, P < 0.01.

Jin et al.





in tumors derived from U87MG overexpressing IL-6, EGFR-vIII, and ID3 were positive for NSC markers (CD15 andNestin), oligodendrocytes (NG2), neurons (Tuj1), and astro-cytes (S100b), as well as endothelial cell markers (CD31),showing tumor cell heterogeneity. In sharp contrast, mostcells comprising tumors derived from ID3-depleted U87MG-EGFRvIII and IL-6–depleted U87MG-ID3 cell were S100bþ

astrocytes, implying that these tumors were homogenous,which is rarely observed in human GBM specimens. Toassess whether the EGFR–ID3–IL-6 signaling axis affectssurvival of brain tumor–bearing animals, we orthotopicallyinjected these cells into nude mice. Overexpression ofEGFRvIII, ID3, and IL-6 significantly shortened the survivalof these tumor-bearing mice, and conversely, depletion ofID3 and IL-6 in the U87MG-EGFvIII and U87MG-ID3, respec-tively, dramatically delayed mice death (Fig. 6C). Thesecombined findings indicate that EGF signaling promotestumor progression and heterogeneity through ID3–IL-6 sig-naling pathway.Taken together, our results suggest that activation of EGF

signaling by genetic alteration (e.g., EGFRvIII) promotesmalig-nant tumor formation and progression through pAKT-pSmad5signaling-driven ID3 induction and subsequent activation ofID3-regulated cytokines, GRO1, IL-6, and IL-8.

Effect of EGFR inhibitors on GSCs with active EGFR-ID3signaling axis

We compared the expression pattern of genes associatedwith the EGFR-ID3 signaling axis in X01 and X03 GSCsduring and after treatment with AG1478, which is an EGFRinhibitor. As shown in Fig. 7A, phosphorylated EGFR, phos-phorylated AKT, p-Smad1,5,8, and ID3 protein levels weremarkedly reduced by AG1478, and withdrawal of this inhib-itor led to reactivation of the signaling pathway. Accordingly,treatment of X01, X03, and U87MG-EGFRvIII cells withAG1478 or gefitinib (another EGFR inhibitor) inhibitedneurosphere formation without any obvious cytotoxiceffects (Supplementary Fig. S7); however, all cells testedseemed to regain colonogenic ability 12 day after removalof EGFR inhibitors (Fig. 7B), suggesting that continuousEGFR inhibition may be required to suppress tumor pro-gression and recurrence.

Discussion

In this study, we show that ID3 overexpression alone con-verts primary Ink4a/Arf�/� astrocytes to cancer cells with GSCfeatures, and EGFR-driven ID3 induction primes GSC genesisby upregulation of IL-6. Previous studies have shown that

Figure 5. Role of ID3- and ID3-regulated cytokines in Notchsignaling and GSC markerexpression. A, real-time RT-PCRanalysis revealed differentialexpression of Notch signaling-related genes (JAG1, HES1, andHEY1) and NSC markers (CD133and Nestin) in the U87MG-control,87MG-ID3, U87MG-GRO1,U87MG-IL-6, and U87MG-IL-8cells. a,b,c,d,e, P < 0.05; A,B,C,D,E,P < 0.01. B, shRNA-mediatedknockdown of ID3 in XO1 and XO2GSC lines suppressed expressionof GSC markers (CD133 andNestin), cytokines (IL-6 and IL-8),and Notch signaling-related genes(Jagged1, Hes1, and Hey1) asassessed by real-time RT-PCR.�, P < 0.05; ��, P < 0.01.

70

60

50

40

30

20

10

0

14

12

10

8

6

4

2

0

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

U87MG cellsABDE ABDE

ABCD

ABC

ABCAB

CE CE CECE CE CE

C

CCEce ce ce

abd abd

Ad

A

a

abBcdE

Control

GRO1

IL6

IL6

IL8

IL8

IL3

U87MG cells

Control

GRO1

IL6

IL8

IL3

JAG1

Jagged1

HES1 HEY1

Hes1 Hey1

XO1-shScr.XO1-shID3XO2-shScr.XO2-shID3

CD133

CD133

NESTIN

Nestin

mR

NA

exp

ress

ion

mR

NA

exp

ress

ion

A

B






constitutively active EGFR signaling confers stemness featuresto breast and lung cancer cells by activating Jagged-Notchsignaling through IL-6–mediated JAK/STAT signaling(24, 26). Similarly, our findings indicate that ID3 switchesglioma cells to GSCs with increased IL-6 and Notch activity,as evidenced by the induction of Jagged1, Hey1 and Hes1;conversely, depletion of ID3 in EGFRvIII-possessing GSCsreduces IL-6, Jagged, Hey1, and Hes1 expression and resultsin a loss of GSC properties. Under in vivo conditions, GSCmaintenance requires specific microenvironments such as aperivascular niche (21). Perivascular nitric oxide or vascularniche factor–pigment epithelium-derived factor playsimportant roles of GSC or NSCs maintenance throughactivation of Notch signal pathway (29, 30). Thus, it is likelythat ID3-regulated cytokines (GRO1, IL-6, and IL-8) play acrucial role in reprogramming of tumor microenvironmentthat is required for maintenance of GSC traits (Fig. 7C).

A recent study has also shown that IL-6 and leukemiainhibitory factor expressed from glioma cell lines and GSCsthat express EGFRvIII enhance EGFR signaling and prolifera-tion of wild-type EGFR-possessing neighboring cells throughgp130–EGFR interaction (31). However, themechanism of howaberrant EGFR signaling activates IL-6 expression is stillunknown. However, our findings indicate that ID3 is a crucialintermediate between RTK and cytokine signaling in GSCbiology and that ID3 and IL-6 may be potential intracellular

and extracellular targets for therapeutic modalities againstCSCs.

Previous studies have reported a plausible mechanismunderlying the regulation of Smad1,5,8 by the EGFR-AKTsignaling pathways. For example, p-AKT inhibits GSK3activity through phosphorylation (32); GSK3-dependentphosphorylation of the Smad1 linker domain is required forpolyubiquitination and subsequent proteasomal degrada-tion of Smad1 (33, 34). However, our data show that EGFRsignaling induces ID3 expression through AKT-dependentinduction of p-Smad5, but not p-Smad1. Furthermore, thisEGFR-AKT–driven phosphorylation of Smad5 is indepen-dent of BMP signaling, which is known to inhibit tumori-genicity of brain tumors (18), because p-Smad1,5,8 levelswere maintained in EGFRvIII-overexpressing cells exposedto the BMP antagonist noggin. Although EGFR activation byoverexpression or mutation is a common oncogenic event inmore than 50% of patients with GBM, only 10% to 20% ofthese patients responded to EGFR inhibitors in a clinicaltrial (35). However, the molecular mechanism underlying theresistance to EGFR inhibitors in GBM remains poorly under-stood. Although pharmaceutical inhibitors of EGFR (AG1478and gefitinib) successfully suppress AKT and Smad1,5,8phosphorylation, ID3 expression, and neurosphere forma-tion in GSCs, these suppressive effects of EGFR inhibitorsrapidly disappear after withdrawal of these inhibitors,

A B

C

Puro-shScr.

IL6-shScr.

EGFRvIIIshScr.

EGFRvIII-shID3

ID3-shScr.

ID3-shIL6

Puro-shScr.

IL6-shScr.

EGFRvIIIshScr.

EGFRvIII-shID3

U87MG-Puro-shScr.

U87MG-Puro-shScr. versus U87MG-IL6-shScr.:

Log-rank P (for significance of difference of survival between groups)

U87MG-Puro-shScr. versus U87MG-EGFRvIII-shScr.:

U87MG-Puro-shScr. versus U87MG-ID3-shScr.:

U87MG-EGFRvIII-shScr. versus U87MG-EGFRvIII-shID3.:

U87MG-ID3-shScr. versus U87MG-ID3-shIL6.:

0.0007

0.0008

0.0006

0.0008

0.0130

U87MG-IL6-shScr.

U87MG-EGFRvIII-shScr.

U87MG-EGFRvIII-shID3

U87MG-ID3-shScr.

U87MG-ID3-shIL6

ID3-shScr.

ID3-shIL6

BCE

AcDF

AbdF

AcDE

7

6

5

4

3

2

1

0

1.0

0.8

0.6

0.4

0.2

0.00 20 40 60 80

Days

U87MG cells

U87MG cells

CD

15+

Nes

tin+

NG

2+S

100β

+Tu

j1+

CD

31+

Tum

or w

eigh

t (g)

Sur

viva

l pro

babi

lity

BCE

BcE

Figure 6. EGFR–ID3–IL-6 signalinggives rise to tumors with cellularheterogeneity. A, representativeimages (top, scale bar, 2 mm) andweight (bottom) of tumors obtainedfrom nude mice (n ¼ 6) injectedsubcutaneously with U87MG-Control, U87MG-ID3, U87MG-EGFRvIII, U87MG-EGFRvIII-shID3,U87MG-IL-6, and U87-ID3-shIL-6cells. a,b,c,d,e,f, P < 0.05; A,B,C,D,E,F,P < 0.01. B, immunofluorescenceimages (200�, scale bar, 50 mm) ofCD15þ NSCs (red) or NG2þ

oligodendrocytes (red); Nestinþ

NSCs (red) or Tuj1þ neurons (red);S100bþ astrocytes (red) in thetumors described in (A).Representativeimmunohistochemistry imagesshowing CD31þ endothelial cells inthese tumors. C, Kaplan–Meiersurvival rates of nude miceorthotopically injected withU87MG-Control, U87MG-ID3,U87MG-EGFRvIII, U87MG-EGFRvIII-shID3, U87MG-IL-6, andU87-ID3-shIL-6 cells.

Jin et al.





resulting in recovery of their initial GSC properties. Thus,our study may provide a plausible explanation for tumorrecurrence that occurs after treatment with gefitinib anderlotinib.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Alexander L. Dent (Bcl6 plasmid) and Akio Soeda (X0GSCs).

Grant Support

This study was supported by the National Research Foundation of Korea(NRF) grant funded by Korea government (MEST; No. 2011-0017544). Y.-W. Sohnis supported by a graduate fellowship program from the Brain Korea 21 projectfunded by MEST. X. Jin is supported by a postdoctoral fellowship program fromthe Brain Korea 21 project funded by MEST. S.-H. Kim is supported by apostdoctoral fellowship program from NRF funded by MEST (NRF-2009-351-C00137).

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 April 23, 2011; revised August 26, 2011; accepted September 20, 2011;published OnlineFirst October 5, 2011.

References1. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med

2008;359:492–507.2. Cancer Genome Atlas Research Network. Comprehensive genomic

characterization defines human glioblastoma genes and core path-ways. Nature 2008;455:1061–8.

3. Libermann TA, Nusbaum HR, Razon N, Kris R, Lax I, Soreq H, et al.Amplification, enhanced expression and possible rearrangement ofEGF receptor gene in primary human brain tumours of glial origin.Nature 1985;313:144–7.

4. Nagane M, Coufal F, Lin H, Boqler O, Cavenee WK, Huang HJ. Acommon mutant epidermal growth factor receptor confers enhancedtumorigenicity on human glioblastoma cells by increasing proliferationand reducing apoptosis. Cancer Res 1996;56:5079–86.

5. Vescovi AL, Galli R, Reynolds BA. Brain tumor stem cells. Nat RevCancer 2006;6:425–36.

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

7. BenezraR,Davis RL, LockshonD, Turner DL,WeintraubH. The proteinId: a negative regulator of helix-loop-helix DNA binding proteins. Cell1990;61:49–59.

8. Alani RM, Young AZ, Shifflett CB. Id1 regulation of cellular senescencethrough transcriptional repression of p16/Ink4a. Proc Natl Acad Sci US A 2001;98:7812–6.

9. Gupta GP, Perk J, Acharyya S, de Candia P, Mittal V, Todorova-Manova K, et al. ID genes mediate tumor reinitiation during breastcancer lungmetastasis. ProcNatl Acad Sci USA2007;104:19506–11.

10. Lyden D, Young AZ, Zaqzaq D, Yan W, Gerald W, O'Reilly R, et al. Id1and Id3 are required for neurogenesis, angiogenesis and vasculariza-tion of tumour xenografts. Nature 1999;401:670–7.

11. Jeon HM, Jin X, Lee JS, Oh SY, Sohn YW, Park HJ, et al. Inhibitor ofdifferentiation 4drivesbrain tumor-initiating cell genesis through cyclinE and notch signaling. Genes Dev 2008;22:2028–33.

12. Anido J, Saez-Borderias A, Gonzalez-Junca A, Rodon L, Folch G,Carmona MA, et al. TGF-b receptor inhibitors target the CD44high/ID1high glioma-initiating cell population in humanglioblastoma. CancerCell 2010;18:655–68.

Figure 7. Effect of EGFR inhibitorsin GSCs with activated EGFR-ID3signaling. A, the EGFR inhibitorAG1478 (10 mmol/L for 12 hrs)markedly reduced levels of ID3,p-AKT, and p-Smad1,5,8 inEGFRvIII-expressing X01 and X03cells, but 12 hours after withdrawal,ID3, p-AKT, and p-Smad1,5,8 wererestored. B, recovery ofneurosphere formation in X01, X03and U87MG-EGFRvIII glioma cellsby withdrawal of EGFR inhibitors.Cells were treated with AG1478(1 mmol/L) and gefitinib (1 mmol/L)for 3 days and then they wereallowed to grow for 12 days withoutAG1478 and gefitinib. a,b,c,d,e,f,P < 0.05; A,B,C,D,E,F, P < 0.01. C,EGFR signaling leads to acquisitionof GSC characteristics andangiogenesis by induction of ID3viathe AKT-dependent activation ofSmad5 and subsequent inductionof ID3-regulated cytokinessignaling.

ID3

ID3

IL8IL6

GRO1

Glioma stem cell genesisangiogenesis

β-Actin

XO1 XO3

XO1 XO3 Puro EGFRvIII

p-Smad1,5,8

p-Smad5

p-AKT

Total-AKT

p-AKT

p-EGFR

EGFR

EGFRvIII

AG1478– + ± – + ±

U87MG cells

100

90

80

70

60

50

40

30

20

10

0

Neu

rosp

here

s

(10 μmol/L, 12 h)

Total-Smad1,5,8

DMSO

AG1488 1 μmol/L

Withdraw AG1488 1 μmol/L

Gefitinib 1 μmol/LWithdraw gefitinib 1 μmol/L

BCD

ABCDe

BcD

ACE

ACE

ACE

BDF

BDFBDF

bd

c cACE ACE

A

B

C






13. Bachoo RM, Maher EA, Ligon KL, Sharpless NE, Chan SS, You MJ,et al. Epidermal growth factor receptor and Ink4a/Arf: convergentmechanisms governing terminal differentiation and transformationalong the neural stem cell to astrocyte axis. Cancer Cell 2002;1:269–77.

14. Soeda A, Park M, Lee D, Mintz A, Androutsellis-Theotokis A, McKayRD, et al. Hypoxia promotes expansion of the CD133-positiveglioma stem cells through activation of HIF-1alpha. Oncogene 2009;28:3949–59.

15. Kim TK, Lee JS, Oh SY, Jin X, Choi YJ, Lee TH, et al. Direct transcrip-tional activation of promyelocytic leukemia protein by IFN regulatoryfactor 3 induces the p53-dependent growth inhibition of cancer cells.Cancer Res 2007;67:11133–40.

16. Korchynskyi O, ten Dijke P. Identification and functional characteriza-tion of distinct critically important bone morphogenetic protein-spe-cific response elements in the Id1 promoter. J Biol Chem 2002;277:4883–91.

17. Hong CC, Yu PB. Applications of small molecule BMP inhibitors inphysiologyanddisease.CytokineGrowthFactor Rev2009;20:409–18.

18. Piccirillo SG, Reynolds BA, Zanetti N, Lamorte G, Binda E, Broggi G,et al. Bone morphogenetic proteins inhibit the tumorigenic potential ofhuman brain tumour-initiating cells. Nature 2006;444:761–5.

19. Brunet LJ,McMahon JA,McMahon AP, Harland RM.Noggin, cartilagemorphogenesis, and joint formation in the mammalian skeleton. Sci-ence 1998;280:1455–7.

20. McReynolds LJ, Gupta S, Figueroa ME, Mullins MC, Evans T. Smad1and Smad5 differentially regulate embryonic hematopoiesis. Blood2007;110:3881–90.

21. Calabrese C, Poppleton H, Kocak M, Hogg TL, Fuller C, Hamner B,et al. A perivascular niche for brain tumor stem cells. Cancer Cell2007;11:69–82.

22. Charles N, Ozawa T, Squatrito M, Bleau AM, Brennan CW, Hambard-zumyan D, et al. Perivascular nitric oxide activates notch signaling andpromotes stem-like character inPDGF-inducedgliomacells. Cell StemCell 2010;6:141–52

23. Benezra R, Rafii S, Lyden D. The Id proteins and angiogenesis.Oncogene 2001;20:8334–41.

24. Sansone P, Storci G, Tavolari S, Guarnieri T, Giovannini C, Taffurelli M,et al. IL-6 triggers malignant features in mammospheres from human

ductal breast carcinoma and normal mammary gland. J Clin Invest2007;117:3988–4002.

25. Ancrile B, Lim KH, Counter CM. Oncogenic Ras-induced secretion ofIL6 is required for tumorigenesis. Genes Dev 2007;21:1714–9.

26. Gao SP, Mark KG, Leslie K, Pao W, Motoi N, Gerald WL, et al.Mutations in the EGFR kinase domain mediate STAT3 activation viaIL-6 production in human lung adenocarcinomas. J Clin Invest2007;117:3846–56.

27. HuangWL,YehHH, LinCC, LaiWW,Chang JY,ChangWT, et al. Signaltransducer and activator of transcription 3 activation up-regulatesinterleukin-6 autocrine production: a biochemical and genetic studyof established cancer cell lines and clinical isolated human cancercells. Mol Cancer 2010;9:1–16.

28. Yu RY, Wang X, Pixley FJ, Yu JJ, Dent AL, Broxmeyer HE, et al. BCL-6negatively regulates macrophage proliferation by suppressing auto-crine IL-6 production. Blood 2005;105:1777–84.

29. Charles N, Ozawa T, Squatrito M, Bleau AM, Brennan CW, Hambard-zumyan D, et al. Perivascular nitric oxide activates notch signaling andpromotes stem-like character inPDGF-inducedgliomacells. Cell StemCell 2010;6:141–52.

30. Andreu-Agullo C, Morante-Redolat JM, Delgado AC, Farinas I. Vas-cular niche factor PEDF modulates Notch-dependent stemness in theadult subependymal zone. Nat Neurosci 2009;12:1514–23.

31. IndaMM,Bonavia R,MukasaA,Narita Y, SahDW,VandenbergS, et al.Tumor heterogeneity is an active process maintained by a mutantEGFR-induced cytokine circuit in glioblastoma. Genes Dev 2010;24:1731–45.

32. Pap M, Cooper GM. Role of glycogen synthase kinase-3 in thephosphatidylinositol 3-Kinase/Akt cell survival pathway. J Biol Chem1998;273:19929–32.

33. Fuentealba LC, Eivers E, Ikeda A, Hurtado C, Kuroda H, Pera EM, et al.Integrating patterning signals: Wnt/GSK3 regulates the duration of theBMP/Smad1 signal. Cell 2007;131:980–93.

34. Sapkota G, Alarcon C, Spagnoli FM, Brivanlou AH, Massague J.Balancing BMP signaling through integrated inputs into the Smad1linker. Mol Cell 2007;25:441–54.

35. Mellinghoff IK, Wang MY, Vivanco I, Haas-Kogan DA, Zhu S, Dia EQ,et al. Molecular determinants of the response of glioblastomas toEGFR kinase inhibitors. N Engl J Med 2005;353:2012–24.

Jin et al.





2011;71:7125-7134. Published OnlineFirst October 5, 2011.Cancer Res Xun Jin, Jinlong Yin, Sung-Hak Kim, et al. Cytokine InductionStem-like Cells and Tumor Angiogenesis by ID3-Driven EGFR-AKT-Smad Signaling Promotes Formation of Glioma

Updated version

10.1158/0008-5472.CAN-11-1330doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2011/10/27/0008-5472.CAN-11-1330.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/71/22/7125.full#ref-list-1

This article cites 35 articles, 11 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/71/22/7125.full#related-urls

This article has been cited by 7 HighWire-hosted articles. Access the articles at:

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

SubscriptionsReprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/71/22/7125To 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-11-1330

http://cancerres.aacrjournals.org/content/suppl/2011/10/27/0008-5472.CAN-11-1330.DC1

http://cancerres.aacrjournals.org/content/71/22/7125.full#ref-list-1

http://cancerres.aacrjournals.org/content/71/22/7125.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/71/22/7125


egfr-akt-smad signaling promotes formation of glioma …preserved id3 expression during noggin...

Documents