Therapeutics, Targets, and Chemical Biology

SPARC Stimulates Neuronal Differentiation ofMedulloblastoma Cells via the Notch1/STAT3 Pathway

Praveen Bhoopathi1, Chandramu Chetty1, Ranadheer Dontula1, Meena Gujrati2, Dzung H. Dinh3,Jasti S. Rao1,3, and Sajani S. Lakka1

AbstractSecreted protein acidic and rich in cysteine (SPARC) participates in the regulation ofmorphogenesis andcellular

differentiation through its modulation of cell–matrix interactions. We previously reported that SPARC expressionsignificantly impairs medulloblastoma tumor growth in vivo. In this study, we show that adenoviral-mediatedoverexpression of SPARC cDNA (Ad-DsRed-SP) elevated the expression of the neuronal markers NeuN, nestin,neurofilament, and MAP-2 in medulloblastoma cells and induced neuron-like differentiation. SPARC overexpres-sion decreased STAT3 phosphorylation; constitutive expression of STAT3 reversed SPARC-mediated expression ofneuronal markers. We also show that Notch signaling is suppressed in the presence of SPARC, as well as the Notcheffectorbasic helix-loop-helix (bHLH) transcription factor hairy andenhancer of split 1 (HES1). Notch signalingwasfound to be responsible for the decreased STAT3 phosphorylation in response to SPARC expression. Furthermore,expression of SPARC decreased the production of interleukin 6 (IL-6) and supplemented IL-6–abrogated, SPARC-mediated suppression of Notch signaling and expression of neuronal markers. Immunohistochemical analysis oftumor sections frommice treatedwithAd-DsRed-SP showed increased immunoreactivity for theneuronalmarkersand a decrease in Notch1 expression and phosphorylation of STAT3. Taken together, our results suggest thatSPARC induces expression of neuronal markers in medulloblastoma cells through its inhibitory effect on IL-6–regulated suppression of Notch pathway–mediated STAT3 signaling, thus giving further support to the potentialuse of SPARC as a therapeutic candidate for medulloblastoma treatment. Findings show that SPARC-inducedneuronal differentiation can sensitizemedulloblastoma cells for therapy.Cancer Res; 71(14); 4908–19.’2011 AACR.

Introduction

Medulloblastoma is a malignant tumor of the cerebellum.The median age at diagnosis is 5 years, with the age rangeextending into young adulthood. Primary management con-sists of surgical resection followed by radiation therapy andchemotherapy. Current therapies have serious short-term andlong-term adverse effects, including neurocognitive deficits,endocrinopathies, sterility, and the risk of developing second-ary high-grade glioma or meningioma (1). Patients withrecurrent disease after primary therapy have a particularlypoor prognosis, with a median survival of less than 6 months;

the 2-year survival rate among these patients is approximately9% (2). Medulloblastomas are functionally heterogeneoustumors with a subset of cells that have a stem-like phenotypethat drives tumor growth. These cells contain 2 karyotypicallysimilar populations—one composed of rapidly proliferating,undifferentiated cells and a second (the nodules) composed oflargely nonproliferative cells that have differentiated intoneurons (3). Tumors with more cells of the second type areknown to be less aggressive clinically (4). This suggests thatmodulating the number of primitive cells in medulloblastomacan affect the clinical outcome.

The Notch signaling pathway regulates cell differentiationby means of intercellular communication between adjacentcells (5). A role for Notch in medulloblastoma stem cellmaintenance has been described. Notch suppression induceddifferentiation in medulloblastoma cells and viable popula-tions of better-differentiated cells continued to grow whenNotch activation was inhibited. However, they were unable toefficiently form soft-agar colonies or tumor xenografts, sug-gesting that a cell fraction required for tumor propagation hadbeen depleted. Furthermore, Notch blockade reduced theCD133-positive cell fraction, suggesting that the loss oftumor-forming capacity could be due to the depletion ofstem-like cells. Stem-like cells in medulloblastoma tumorsthus seem to be selectively vulnerable to agents inhibiting theNotch pathway (6).

Authors' Affiliations: 1Program of Cancer Biology, Department of CancerBiology and Pharmacology, 2Department of Pathology, and 3Departmentof Neurosurgery, University of Illinois College ofMedicine at Peoria, Peoria,Illinois

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

P. Bhoopathi and C. Chetty contributed equally.

Corresponding Author: Sajani S. Lakka, Department of Cancer Biologyand Pharmacology, University of Illinois College of Medicine at Peoria, OneIllini Drive, Peoria, IL 61605. Phone: 309-671-3445; Fax: 309-671-3442;E-mail: slakka@uic.edu

doi: 10.1158/0008-5472.CAN-10-3395

’2011 American Association for Cancer Research.

CancerResearch

Cancer Res; 71(14) July 15, 20114908

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

Secreted protein acidic and rich in cysteine (SPARC/osteo-nectin/BM40) is a 32-kDa calcium-binding glycoprotein thataffects a wide variety of cellular processes, including counter-adhesion, extracellular matrix remodeling (7), cell migration(8), and angiogenesis (7, 9). In adult tissues, expression ofSPARC is highest in tissues undergoing active matrix remo-deling, where it is hypothesized to function as a negativeregulator of growth factor activity and has been shown topossess antiadhesive and antiproliferative properties (10–12).SPARC is also thought to play a critical role in epithelialdifferentiation (13), parietal endoderm differentiation, cardi-omyogenesis in embryoid bodies (14), and differentiation ofmyoblasts (15). In addition, SPARC was shown to play a role inneural and/or glial differentiation (16). Furthermore, previousstudies have shown that Daoy and D283medulloblastoma cellsare arrested along the neuronal differentiation pathway (17).In a recent study, we established that SPARC expression

induced apoptosis and regressed preestablished tumor growthin medulloblastoma cells (18). In the present study, we firstinvestigated the effect of endogenous expression of SPARC byusing adenoviral-mediated delivery of SPARC full-lengthcDNA on the expression of neuronal markers in medulloblas-toma cells in vitro and in vivo. Furthermore, we show the roleof Notch1/STAT3 in SPARC-induced neuronal marker expres-sion in human medulloblastoma cell lines. These data define anew role for SPARC as an inducer of neuronal differentiationand could lead to a favorable course of treatment of medullo-blastoma patients and it may sensitize tumors for therapy.

Materials and Methods

Cell culturesWe used the Daoy (HTB-186), D283Med (HTB-185), UW228,

D425 cell lines and primary medulloblastoma cells (H2405 andH2411) for this study. Daoy and D283 cells were authenticatedby DNA profile, using the short tandem repeat (STR), cyto-genetic analysis, and isoenzymes and obtained from AmericanType Culture Collection (ATCC). D425, H2405, and H2411 cellswere kindly provided by Dr. Darell D. Bigner (Duke UniversityMedical Center); UW228 cells were kindly provided by Dr Ali-Osman (Duke University Medical Center) in June 2010. Thesecells were authenticated on the basis of c-Myc amplificationand chromosomal aberrations by the provider (19, 20). At thethird or fourth passage, cells were frozen and these frozenstocks were used for further experimental studies up to thetenth passage to obtain consistent results. Daoy cells werecultured in Advanced-MEM, and D283, D425, H2405, andH2411 were cultured in Improved-MEM (Zn Option) andUW228 cells were cultured in RPMI-1640. All of these mediawere supplemented with 10% FBS, 2 mmol/L L-glutamine, 2mmol/L sodiumpyruvate, 100 units/mL penicillin, and 100 mg/mL streptomycin. Cells were maintained in a humidifiedatmosphere containing 5% CO2 at 37�C.

Antibodies and reagentsAntibodies against SPARC, STAT3, phosphorylated STAT3

(pSTAT3; Tyr705), pSTAT3 (Ser727), Notch1, Notch2, Notch3,hairy and enhancer of split 1 (HES1), glyceraldehyde-3-phos-

phate dehydrogenase (GAPDH; Santa Cruz Biotechnology),nestin, mitogen-activated protein 2 (MAP-2), neurofilament(NF; Novus Biologicals), NeuN (Millipore), NeuN (InvitrogenCorporation), MAP-2 (Bethyl Laboratories, Inc.), neutralizinginterleukin 6 (IL-6) antibody (Biovision), Calcium Green-2-AM(Invitrogen), and N-[(3,5-difluorophenyl)acetyl]-L-alanyl-2-phenyl]glycine-1,1-dimethylethyl ester (DAPT; Sigma), paraf-fin-embedded humanmedulloblastoma tumor sections (a kindgift from Dr. Dan Fults, Comprehensive Care for Brain andSpine Disorders, Salt Lake City, UT) were used in this study. Allother reagents were of analytic reagent grade or better.

Adenovirus construction and adenoviral infectionWe constructed adenoviral vectors carrying full-length

human SPARC cDNA (Ad-DsRed-SP) and an empty vector(Ad-DsRed) using Adeno-X ViraTrak Expression System-2(Clontech Laboratories) as described previously (21). Thegeneration, amplification, and titer of the adenovirus wereconducted according to previously described procedures (21).Infection with recombinant viruses was accomplished byexposing cells to adenovirus in serum-free cell culture med-ium for 1 hour followed by the addition of serum-containingmedium. Cells were then incubated for various time periods asdetailed in the following experiments.

ImmunoblottingImmunoblot analysiswas conducted as described previously

(22). Briefly, cells were infected with mock, 100 multiplicity ofinfection (MOI) of Ad-DsRed, or various MOI of Ad-DsRed-SPand incubated for 36 hours at 37�C. Cell lysates were preparedand equal amounts of protein was resolved on SDS-PAGE geland transferred on to polyvinylidene difluoride membrane.Next, the blot was blocked and probed overnight with primaryantibody at 4�C followed by horseradish peroxidase (HRP)-conjugated secondary antibody for 1 hour, and signals weredetected using enhanced chemiluminescence reagent.

Transfection with plasmidsPlasmid-expressing constitutively active STAT3 (pSTAT3-

C) was obtained from Addgene Inc. (plasmid 8722). Plasmidsexpressing IL-6 and HES1 were obtained from Origene Tech-nologies. All transfection experiments were conducted withFuGeneHD (Roche) transfection reagent according to themanufacturer's protocol.

Intracellular Ca2þ monitoringDaoy/D283 cell were plated in 96-well plates and infected

with mock, 100 MOI of Ad-DsRed or 100 MOI of Ad-DsRed-SPfor 36 hours andmeasured intracellular Ca2þ concentration asdescribed previously (23). The Calcium Green-2-AM fluores-cence was expressed as (Fmax� Fmin)/Fmin, where Fmax was themaximum and Fmin the minimum fluorescence measured ineach well.

Immunofluorescence and immunohistochemicalanalyses

We used a previously described protocol with minorchanges (24). Briefly, the cells were infected with adenovirus

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4909

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

as above for 36 hours. Cells were fixed, permeabilized, blocked,and incubated with primary antibody overnight at 4�C, fol-lowed by fluorescein isothiocyanate–conjugated secondaryantibody for 1 hour. Finally, cells were mounted with Vecta-shield mounting medium (Vector Labs). The results weredocumented using fluorescence microscope. For immunohis-tochemical analysis, tissue sections (4–5 mm thick) weredeparaffinized, rehydrated, washed with PBS, permeabilized,and incubated overnight with primary antibodies. Tissuesections were then incubated with HRP-conjugated secondaryantibodies, followed by 3,30-diaminobenzidine peroxidase sub-strate; Sigma) solution, and then counterstained with hema-toxylin andmounted. The images were processed as describedpreviously (24).

Intracranial tumor modelThe animal experiments were carried out as described

previously by us (24). D425 (1 � 105 cells/10 mL PBS) cellswere stereotactically implanted. After 14 days of tumor cellimplantation, the animals were randomized into 3 groups(10 mice/group). Each mouse received 3 intratumoral injec-tions on days 15, 17, and 19 with PBS, Ad-DsRed [5 � 107

plaque forming units (PFU)] or Ad-DsRed-SP (5 � 107 PFU)in 10 mL of volume. Animals were monitored for up to 90days, which is when we arbitrarily terminated the experi-ment. Mice brains were fixed in 10% buffered formalinand embedded in paraffin. Tissue sections (5 mm thick)were obtained from the paraffin blocks and stained withhematoxylin and eosin (H&E) by using standard histologictechniques. Tissue sections were also subjected to immu-nostaining as described above.

Statistical analysisAll the data are expressed as mean� SD. Statistical analysis

was conducted using the Student's t test or a one-way ANOVA.P < 0.05 was considered significant.

Results

SPARC induces neuronal differentiation ofmedulloblastoma cells

We observed very low or minimal staining for SPARC inhuman medulloblastoma tissue samples compared with nor-mal cerebellum (Fig. 1A). Dual immunoassay of these tissuesamples for neuronal markers and SPARC indicated that veryfew cells stained positive for neuronal makers and thatSPARC-expressing tumor cells stained positive for NeuNand nestin neuronal markers (Fig. 1B and C). Furthermore,previous studies have shown that Daoy and D283 medullo-blastoma cells are arrested along the neuronal differentiationpathway (17). We therefore determined weather SPARCinduced the expression of neuronal markers in Daoy, D283,UW228, and D425 medulloblastoma cell lines as well as H2405and H2411 primary medulloblastoma cells in vitro and in vivo.To examine the effect of SPARC expression on neuronaldifferentiation, we used an adenoviral vector–expressingSPARC cDNA (Ad-DsRed-SP) in above cell lines. Medulloblas-toma cells infected with Ad-DsRed-SP showed obvious neurite

extensions and classic neuronal features, such as shrinkage ofthe cell body, and these cells were distinguished by highlyrefractive cell bodies with neuron-like processes terminatingin structures that resembled growth cones (Fig. 2A; Supple-mentary Fig. S1). In contrast, few if any empty vector (Ad-DsRed)-infected cells differentiated into neurons under thesame conditions (Fig. 2A). These findings suggest that themedulloblastoma cells treated with Ad-DsRed-SP might havedifferentiated into neuron-like cells. To further characterizethis neuronal induction phenomenon, we examined neuronalmarker expression by immunoblot and immunocytochemicalanalyses. The cells that exhibited contracted cell bodies andprocesses showed clear staining for the neuronal-specificmarkers NeuN, nestin, MAP-2, and NF. In contrast, the flat,unresponsive medulloblastoma cells remained unstained(Fig. 2B). The percentage of cells positively reactive for NeuN,nestin, MAP-2, and NF was 52% � 2.5%, 55% � 2.3%, 56% �3.8%, and 53% � 2.4%, respectively. Immunoblotting wasconducted to confirm these results, and the protein analysisshowed a marked correlation with the results of the immu-nocytochemical analysis (Fig. 2C; Supplementary Fig. S2).SPARC expression induced a statistically significant increase(P < 0.05) of up to 4-fold in the protein expression levels of allthe markers as compared with empty vector-treated cells inmedulloblastoma cell lines. Similar to adenoviral-mediatedexpression of SPARC cDNA, recombinant human SPARCprotein (rhSPARC) induced neuron-like cell morphologyand expression of neuronal markers in medulloblastomacell lines (Supplementary Fig. S3).

Activity-dependent changes in neuronal processes such assynaptic plasticity and neuronal survival are mediated in largepart through elevations in intracellular calcium levels. In manycases, this involves stimulation of calcium influx, particularlyvia voltage-operated L-type channels or receptor-operated N-methyl-D-aspartate channels (25). To test the neuronal natureof the differentiated cells from Daoy/D283 treated with Ad-DsRed-SP, we examined the changes in intracellular free Ca2þ

concentration, [Ca2þ]i, in response to depolarization of mem-brane potential with superfusion of high KCl (100 mmol/L),using Calcium Green-2-AM. Figure 3A shows that intensefluorescence in Ad-DsRed-SP-infected cells compared withmock and Ad-DsRed–infected Daoy/D283 cells. We next con-ducted dual-wavelength excitation fluorimetry using in Cal-cium Green-2-AM–loaded cells. High KCl bathing solution for60 seconds increased [Ca2þ]i in differentiated neuronal cells(Fig. 3B). In contrast, cells treated with mock and Ad-DsRedhad no changes in [Ca2þ]i by membrane depolarization withhigh KCl.

SPARC expression decreases STAT3 signalingSTAT3 signaling has been shown to regulate cell fate

determination, renewal, and differentiation in various cells(26). We evaluated whether suppression of STAT3 cell signal-ing machinery could modulate the neuronal fate of SPARC-expressing medulloblastoma cells. To elucidate the effects ofSPARC overexpression on STAT3 signaling, cells were infectedwith Ad-DsRed-SP for 36 hours and homogenized in lysisbuffer. Then, lysates were immunoblotted with antibodies

Bhoopathi et al.

Cancer Res; 71(14) July 15, 2011 Cancer Research4910

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

against pSTAT3. Densitometric analysis revealed that theSTAT3 phosphorylation was reduced significantly (70.5% inDaoy cells and 67% in D283 cells) as compared with the mockor empty vector-infected cells (Fig. 4A). Furthermore, SPARC

expression in primary medulloblastoma cells (H2405 andH2411), D425, and UW228 cells also suppressed STAT3 phos-phorylation (Supplementary Fig. S2). We therefore speculatedthat suppression of STAT3 influenced the differentiation fate

Figure 1. A, SPARC expression inhuman medulloblastoma tissue.Human medulloblastoma tissuesections (samples 1–12) weresubjected to immunohistochemicalanalysis using SPARC-specificantibody. SPARC expression ishigh in normal brain tissue (20�). Incontrast, the SPARC expression isvery weak in most of themedulloblastoma samples (20�).Also shown is a negative controlusing nonspecific IgG. B and C,association of SPARC expressionwith neuronal differentiation inhuman medulloblastoma tumorsections. Paraffin-embeddedtissue sections from humanmedulloblastoma tumorswere analyzed by dualimmunofluorescence for SPARC(green) and neuronal makers NeuNor nestin (red) expression by usingspecific antibodies (10�). Enlargedportions of the sections indicatedwith yellow square boxes (60�).Very few cells express neuronalmarkers that also stain positive forSPARC expression (yellow).

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4911

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

of neural stem cells. To confirm this notion, we transfectedcells with plasmid-expressing pSTAT3-C along with Ad-DsRed-SP infection and total cell lysates. After inoculation,cells were homogenized in lysis buffer and immunoblottedwith antibodies against MAP-2, NF, NeuN, and nestin. Asillustrated in Figure 4B, transfection with plasmid-expressingSTAT3-C induced STAT3 phosphorylation in Ad-DsRed-SP–treated cells comparable with that of the mock and emptyvector-transfected cells. Densitometric analysis indicated thatpSTAT3-C transfection reversed SPARC-induced MAP-2, NF,NeuN, and nestin neuronal markers by up to 50% (P < 0.05) inAd-DsRed-SP–infected cells. Furthermore, STAT3 siRNA

transfection decreased STAT3 levels in Daoy/D283 cells andinduced neuronal markers (Supplementary Fig. S4). Theseresults clearly indicate the role of STAT3 inhibition in thedifferentiation of medulloblastoma cells. Taken together,these data suggested that STAT3 regulates the expressionof neuronal markers.

SPARC induces expression of neuronal markers byblocking Notch signaling

Blockade of Notch signaling, a condition known to induceneuronal differentiation, represses inhibitory basic helix-loop-helix transcription factor-1, HES1 expression (27). It was shown

DaoyA

B

C

Mock

MA

P-2

NF

Neu

NN

esti

n

Ad-DsRed Ad-DsRed-SP

Ad-DsRed-SP

SPARC

NeuN

Nestin

NF

MAP2

SPA

RC

Neu

N

Nes

tin

NF

MA

P2

GAPDH

Ad-DsRed-SP

Ad

-DsR

ed

50 M

OI

100

MO

I

Mo

ck

Ad

-DsR

ed

50 M

OI

100

MO

I

Mo

ck

Ad-DsRed50 MOI 100 MOIMock

5

4

3

3

2

2

1

1

0

0

4

Rel

ativ

e p

rote

in e

xpre

ssio

n

Mock Ad-DsRed Ad-DsRed-SP

D283

Daoy D283

Daoy

D283

Figure 2. Overexpression ofSPARC in medulloblastoma cellsinduces expression of neuronalmarkers. DAOY and D283 cellswere infected with mock, 100 MOIAd-DsRed, or the indicated MOI ofAd-DsRed-SP for 36 hours. A,morphology of medulloblastomatumor cells treated with 100 MOIof Ad-DsRed-SP, using phase-contrast microscopy. B,immunocytochemical analysis forneuronal markers inmedulloblastoma cells treatedwith 100 MOI of Ad-DsRed-SP. C,cell lysates were assessed forSPARC and neuronal markers byimmunoblot analysis by usinganti-SPARC, anti-NeuN, anti-nestin, anti-NF, and anti-MAP-2antibodies. Glyceraldehyde-3-phosphate dehydrogenase(GAPDH) served as a control.Results are representative of 3independent experiments. Proteinband intensities were quantifiedby using ImageJ software (NIH).Columns, mean of 3 experiments;bars, SD; *, P < 0.01, significantdifference from Ad-DsRed.

Bhoopathi et al.

Cancer Res; 71(14) July 15, 2011 Cancer Research4912

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

that STAT3 is activated in the presence of active Notch, as wellas the Notch effectors HES1 and HES5 (28). We thereforeexamined the effects of Ad-DsRed-SP infection on the expres-sion of Notch family members (Notch1 and 2) and HES1 inmedulloblastoma cells. Immunoblot analysis revealed thatNotch1 and Notch2 were suppressed in a dose-dependentmanner reaching about 75% inhibition in Daoy and 70%inhibition in D283 cells at 100 MOI Ad-DsRed-SP infectionwhen compared with mock or Ad-DsRed–infected cells(Fig. 5A). Similarly, HES1 was inhibited by 68% in Daoy and65% in D283 (P < 0.05) cells infected with Ad-DsRed-SP ascompared withmock or Ad-DsRed controls (Fig. 5A). Similarly,SPARC overexpression in the primary medulloblastoma cells(H2405 and H2411), D425, and UW228 cell lines lead todecreased Notch1 and HES1 expression (SupplementaryFig. S2). Because we observed that STAT3 inhibition contri-butes to the expression of neuronal markers in SPARC-over-expressed cells, we therefore used HES1 cDNA to determinewhether HES1 expression is necessary for STAT3-mediatedinduction of neuronal markers in SPARC-overexpressed cells.Transfection ofmedulloblastoma cellswith a vector specific forHES1 cDNA in SPARC-overexpressed cells resulted in anincrease in the abundance of HES1 protein comparable withmock or Ad-DsRed–treated cells (Fig. 5B). Concomitantly,densitometric analysis revealed that STAT3 phosphorylationwas increased significantly by 70% and 68% (P < 0.05 vs. Ad-DsRed-SP) in Daoy/D283 cells (Fig. 5B). Furthermore, neuron-like morphologic changes and the induction of neuronalmarkers as determined by immunocytochemical analysis

and immunoblotting, respectively, were suppressed by HES1overexpression in SPARC-overexpressed cells (Fig. 5B, Supple-mentary Fig. S1). Together, these results suggest thatHES1 is anessential mediator of the action of STAT3 in SPARC-inducedneuronal differentiation in medulloblastoma cells.

Effects of SPARC siRNA on Notch expressionTo confirm that SPARC can induce neurogenesis in medul-

loblastoma cells viaNotch1-mediatedHES1 signaling,we exam-ined the effects of SPARCsiRNA (SP-siRNA)on the expressionofNotch family members and neuronal markers in medulloblas-toma cells. Figure 5C indicates that infectionwith an adenoviralvector encodingSP-siRNAdecreased SPARC levels as comparedwith mock or control siRNA–treated cells. Along with SPARCreduction, therewas induction ofNotch1,HES1 expression, andSTAT3 phosphorylation and suppression of the expression ofNeuN and MAP-2 neuronal markers in SP-siRNA–treated cells(Fig. 5C). Blocking Notch1 using a known g-secretase inhibitorDAPT (29) in SP-siRNA–treated cells suppressed HES1 andSTAT3 phosphorylation and induced the expression of neuro-nalmarkers (Fig. 5C). Taken together, these results suggest thatSPARC-induced neuronal differentiation by blocking Notch-mediated STAT3 phosphorylation.

IL-6 regulates Notch-mediated modulation of neuronalmarkers in SPARC-overexpressed medulloblastomacells

Previous studies show that SPARC expression attenuatedIL-6 secretion (30) and that IL-6 upregulates Notch signaling

Figure 3. KCl-induced Ca2þ

imaging in live differentiatedmedulloblastoma cells. A,medulloblastoma cells wereplated in 8-well chamber slidesand infected with mock, Ad-DsRed, or Ad-DsRed-SP for 36hours. The Calcium Green-2-AMand KCl was added andphotographed using afluorescence microscope. B,medulloblastoma cells wereplated in 96-well plates, infectedwith mock, Ad-DsRed, or Ad-DsRed-SP for 36 hours, andassessed for the changes inintracellular free Ca2þ

concentration. Representativetime–response data for KClexcitation using differentiatedneuron-like cells. Cells weretreated with Calcium Green-2-AMin the presence and absence ofKCl, and kinetic fluorometricreading was obtained using FlexStation and the representativegraph is shown. Results arerepresentative of 3 independentexperiments. Symbols, mean of 3experiments; bars, SD.

Mock

0.6

0.4

Flu

ore

scen

ce u

nit

s

0.2

010 20 30 40

Time (s)50 60 70 80 90

D28

3D

aoy

Ad-DsRed Ad-DsRed-SP

Daoy mockDaoy Ad-DsRedDaoy Ad-DsRed-SPD283 mockD283 Ad-DsRedD283 Ad-DsRed-SP

Mock Ad-DsRed

+100 mmol/L KCIA

B

Ad-DsRed-SP

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4913

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

(31). Therefore, we examined the role of IL-6 in SPARC-induced Notch signaling and expression of neuronalmarkers. Immunoblot analysis for IL-6 expression indicatesthat SPARC overexpression decreased IL-6 in a dose-depen-dent manner in Daoy, D283, D425, and UW228 cell lines andprimary medulloblastoma cells (H2405 and H2411; Fig. 6A;Supplementary Fig. S2). To better understand the role of IL-

6–mediated effects on neuronal markers in SPARC-expressed cells, we overexpressed IL-6 in SPARC-overex-pressed medulloblastoma cells. Figure 6B indicated thatNotch1 expression increased by 65% and 60% in IL-6 andSPARC-overexpressed cells as compared with only SPARC-overexpressed medulloblastoma cells. Consequently, theexpression of neuronal markers MAP-2, NF, and NeuN

Daoy

Ad

-DsR

ed

Ad

-DsR

ed-S

P

Ad

-DsR

ed-S

P +

pS

TAT

3-C

Ad

-DsR

ed+p

EV

pS

TAT

3-C

Ad

-DsR

ed

Ad

-DsR

ed-S

P

Ad

-DsR

ed-S

P +

pS

TAT

3-C

Ad

-DsR

ed+p

EV

pS

TAT

3-C

Ad-DsRed-SP

Mock50 MOI

pSTAT3-Y-705

Ad-DsRed

Ad-DsRed-SPpStat3-C

pS

TAT

3

Neu

N

MA

P-2

Ad-DsRed + pEVAd-DsRed-SP + pSTAT3-C

pSTAT3-S-727

Daoy

Daoy

D283

D283

pSTAT3-Y-705

pSTAT3-S-727

**

*

*

*

*

*

*

** **

** ** **

**

**

Rel

ativ

e p

rote

in e

xpre

ssio

nR

elat

ive

pro

tein

exp

ress

ion

Ad-DsRed100 MOI

120

80

40

0

300

150

0

300

150

0

STAT3

pSTAT3(Tyr705)

pSTAT3(Tyr727)

GAPDH

STAT3

SPARC

pSTAT3(Tyr705)

NF

MAP-2

Nestin

NeuN

GAPDH

Ad-DsRed-SP

A

B

Ad

-DsR

ed

50 M

OI

100

MO

I

Mo

ck

Ad

-DsR

ed

50 M

OI

100

MO

I

Mo

ck

D283

Figure 4. SPARC expression decreased STAT3 signaling. A, medulloblastoma cells were infected with mock, 100 MOI Ad-DsRed, or the indicated MOIof Ad-DsRed-SP for 36 hours, and cell lysates were assessed for STAT3 and pSTAT3 by using specific antibodies. GAPDH served as a control. Protein bandintensities were quantified by using ImageJ software (NIH). Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference from Ad-DsRed. B,cells were transfected with plasmid constitutively expressing active-STAT3 (pSTAT3-C) for 16 hours and infected with mock, 100 MOI Ad-DsRed,or Ad-DsRed-SP for an additional 24 hours. Cell lysates were assessed for SPARC, activation of STAT3, and neuronal markers by using their specificantibodies. Results are representative of 3 independent experiments. Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference fromAd-DsRed; **, P < 0.05, significant difference from Ad-DsRed-SP.

Bhoopathi et al.

Cancer Res; 71(14) July 15, 2011 Cancer Research4914

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

was also decreased by about 65%, 63%, and 60%, respectively,in cells overexpressed with IL-6 and SPARC compared withonly SPARC-overexpressed cells (Fig. 6B; P < 0.05), suggest-ing that SPARC-mediated suppression of IL-6 leads to

suppression of Notch1 expression, which, in turn, leads toinduction of neuronal markers.

To confirm the role of IL-6 in Notch signaling–mediatedexpression of neuronal markers in SPARC-expressed cells, we

A

B

C

Ad-DsRed-SP Ad-DsRed-SP

SPARCM

ock

Mock

Daoy

Daoy

Daoy

Rel

ativ

e p

rote

in e

xpre

ssio

nR

elat

ive

pro

tein

exp

ress

ion

Rel

ativ

e p

rote

in e

xpre

ssio

n

20

80

40

0

300

200

100

0

300

200

100

0300

200

100

0

300

200

100

0

Not

ch1

Not

ch2

Not

ch1

HE

S1

HE

S1

Not

ch2

D283

D283

10 MOI50 MOI

*

*

* *

***

*

*

* *

**

*

******

** ** **

** **

** ** ** **

****

* * ** *

Ad-DsRed25 MOI100 MOI

Mock

HES1

Notch1 NeuND283

MAP-2

NeuN NFMAP-2

Ad-DsRed-sPpHES1

Mock Control siRNADAPTSP-siRNA

DAPT+Ad-SP-siRNA

Ad-DsRedpEVAd-DsRed-SP+pHES

10 M

OI

25 M

OI

50 M

OI

100

MO

I

Ad-

DsR

ed

Moc

kM

ock

SP

-siR

NA

DA

PT

DA

PT

+S

P-s

iRN

A

Con

trol

siR

NA

Moc

k

SP

-siR

NA

DA

PT

DA

PT

+S

P-s

iRN

A

Con

trol

siR

NA

Ad-

DsR

ed-S

PpE

VpH

ES

1A

d-D

sRed

-SP

+pH

ES

1

Ad-

DsR

ed

Moc

k

Ad-

DsR

ed-S

PpE

VpH

ES

1A

d-D

sRed

-SP

+pH

ES

1

Ad-

DsR

ed

Moc

k

10 M

OI

25 M

OI

50 M

OI

100

MO

I

Ad-

DsR

ed

Notch1

Notch2

Notch3

HES1

GAPDH

GAPDH

Daoy D283

Daoy D283

NFMAP-2NeuNpSTAT3Notch1HES1

GAPDH

MAP-2

NeuN

pSTAT3

Notch1

SPARC

HES1

Figure 5. SPARC inhibits Notch signaling and induces expression of neuronal markers. A, medulloblastoma cells were infected with mock, 100 MOIAd-DsRed, or the indicated MOI of Ad-DsRed-SP for 36 hours, and cell lysates were assessed for Notch family molecules (Notch1, Notch2, and Notch3)and HES1 by using molecule-specific antibodies. GAPDH served as a control. Protein band intensities were quantified by using ImageJ software (NIH).Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference from Ad-DsRed. B, cells were transfected with plasmid-expressing HES1 for 16hours and infected with mock, 100 MOI Ad-DsRed, or Ad-DsRed-SP for an additional 24 hours. Cell lysates were assessed for neuronal markers byimmunoblotting by using their specific antibodies. Results are representative of 3 independent experiments. GAPDH served as a control. Columns, mean of 3experiments; bars, SD; *, P < 0.01, significant difference from Ad-DsRed; **, P < 0.05, significant difference from Ad-DsRed-SP. C, cells were infected withmock, 100 MOI of control siRNA or SP-siRNA for 24 hours. Cells were then treated with 10 mmol/L DAPT and cells were allowed to grow for an additional 12hours. Cell lysates were assessed for Notch signaling molecules and neuronal markers by immunoblotting using their specific antibodies. Results arerepresentative of 3 independent experiments. GAPDH served as a control. Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference fromcontrol siRNA; **, P < 0.01, significant difference from SP-siRNA.

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4915

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

conducted parallel experiments using SP-siRNA. Figure 6Cindicates that SP-siRNA suppressed SPARC levels when com-pared with mock or control siRNA–treated cells. Suppressionof SPARC using SP-siRNA–induced IL-6 and Notch1 expres-

sion by 2- to 3-fold compared with control siRNA–treatedcells. Consequently, the expression of neuronal markers MAP-2, NeuN, and NF was decreased by 60% to 70% in SP-siRNA–treated cells. To further confirm that IL-6 mediates Notch1

A

B

C

Ad-DsRed-SP

Daoy

Mo

ck

50 M

OI

100

MO

I

Ad

-DsR

edM

ock

Co

ntr

ol s

iRN

A+I

L-6

Ab

Co

ntr

ol s

iRN

A

SP

-siR

NA

SP

-siR

NA

+IL

-6 A

b

Mo

ck

Co

ntr

ol s

iRN

A+I

L-6

Ab

Co

ntr

ol s

iRN

A

SP

-siR

NA

SP

-siR

NA

+IL

-6 A

b

Mo

ck

50 M

OI

100

MO

I

Ad

-DsR

ed

Mo

ck

50 M

OI

100

MO

I

Ad

-DsR

ed

Mo

ck

50 M

OI

100

MO

I

Ad

-DsR

ed

D283

SPARC

120

80

40

0

300

200

100

0

300

200

100

0

300

200

100

0

300

200

100

0

pSTAT3

Notch1

GAPDH

– – – – + + + +– – + + – – + +

– – – – + + + +– – + + – – + + Ad-DsRed+SP

IL-6

IL-6

Mo

ck

Ad

-DsR

ed

100

MO

I

50 M

OI

Mo

ck

Ad

-DsR

ed

100

MO

I

50 M

OI

Ad-DsRed-SP

Mock50 MOI

Daoy

Daoy

Daoy

Daoy

D283D283

Daoy D283 D283

Rel

ativ

e p

rote

in le

vels

Rel

ativ

e p

rote

in e

xpre

ssio

nR

elat

ive

pro

tein

exp

ress

ion

IL-6

Mock

IL-6

IL-6

Notch1

HES1

pSTAT3

NF

NeuN

MAP-2

GAPDH

IL-6

Notch1

SPARC

NF

NeuN

MAP-2

GAPDH

Notch1

MockControl siRNA-IL-6 AbSP-siRNA+IL-6 Ab

Notch1 MAP-2 NeuN NF

Control siRNA

SP-siRNA

MAP-2NeuNNF

Ad-DsRed-SP-100 MOIAd-DsRedAd-DsRed+pIL-6

Ad-DsRed-SP-100 MOI+pIL-6

Notch1

* *

*

* *

**

*

** ** ** ** **

*

* **

** ** ****

**

* *IL-6 Notch1

D283

Ad-DsRed100 MOI

Figure 6. SPARC overexpression in medulloblastoma cells inhibits IL-6 signaling and induces expression of neuronal markers. A, medulloblastoma cells wereinfected with mock, 100 MOI Ad-DsRed, or the indicated MOI of Ad-DsRed-SP for 36 hours. Cell lysates were assessed for SPARC, IL-6, pSTAT3, andNotch1 by immunoblot analysis by using specific antibodies. GAPDH served as a control. Protein band intensities were quantified by using ImageJsoftware (NIH). Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference from Ad-DsRed. B, cells were transfected with a plasmid-expressing IL-6 for 16 hours and infected with mock, 100 MOI Ad-DsRed, or the indicated MOI of Ad-DsRed-SP for an additional 24 hours. Cell lysateswere assessed for neuronal markers by using their specific antibodies. Results are representative of 3 independent experiments. Protein band intensitieswere quantified using ImageJ software (NIH). Columns, mean of 3 experiments; bars, SD; *, P < 0.01, significant difference from Ad-DsRed; **, P < 0.05,significant difference from Ad-DsRed-SP. C, cells were infected with mock, 100 MOI control siRNA, or SP-siRNA for 24 hours and treated with IL-6–blockingantibody (Ab) for further 12 hours. Cell lysates were assessed for IL-6, Notch1, and neuronal markers by using their specific antibodies. Results arerepresentative of 3 independent experiments. Protein band intensities were quantified by using ImageJ software (NIH). Columns, mean of 3 experiments; bars,SD; l, P < 0.01 significant difference from control siRNA; @, P < 0.01 significant difference from SP-siRNA.

Bhoopathi et al.

Cancer Res; 71(14) July 15, 2011 Cancer Research4916

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

suppression, which, in turn, regulates the expression of neu-ronal markers in SPARC-modulated cells, we blocked IL-6activity by using neutralizing antibodies and determined thelevels of Notch1 and neuronal markers. IL-6–neutralizingantibody suppressed Notch1 and induced the expression ofneuronal markers in SP-siRNA–treated cells (Fig. 6C), suggest-ing that SPARC-mediated effects on Notch1 regulation aremediated by IL-6.In summary, these results suggest that the inhibitory effect

of SPARC on IL-6 leads to Notch-mediated expression ofneuronal markers in SPARC-overexpressed cells.

SPARC-mediated expression of neuronal markersin vivoWe have previously shown that SPARC expression inhibits

medulloblastoma tumor growth in vivo in an intracranialmodel (18). The present findings raise the question of whetherthe effects of SPARC on tumor growth inhibition are related tothe effect of SPARC on neuronal differentiation in vivo. Immu-nohistochemical analysis was conducted on establishedtumors frommice implantedwith D425medulloblastoma cellsand treated with mock, Ad-DsRed, or Ad-DsRed-SP with anti-bodies specific to detect neuronal markers of human origin.The results show a clear increase in the expression of neuronalmarkers MAP-2, NeuN, nestin, and NF in tumor sections fromAd-DsRed-SP–treated mice as compared with sections frommock and Ad-DsRed–treated animals, thereby suggesting thatSPARC expression induced the expression of neuronalmarkersin vivo (Fig. 7). To determine whether SPARC regulates Notchand STAT3 phosphorylation in vivo, phosphorylation of STAT3and Notch1 expression was measured by immunohistochem-ical analysis. Consistent with the in vitro results, a decrease incleaved Notch1 and phosphorylation of STAT3 was found inAd-DsRed-SP–treated tumors (Fig. 7).

Discussion

Medulloblastoma show a tremendous clinical heterogene-ity, and the degree of neuronal tumor cell differentiationinfluences patient outcome (4). Several studies show thatSPARC induces differentiation; however, no studies haveshown the functional mechanism by which SPARC inducesneuronal differentiation in tumor cells. In the present study,we show that SPARC expression, using an adenoviral vector–expressing SPARC cDNA (Ad-DsRed-SP), induced neuronaldifferentiation in medulloblastoma tumor cells. Furthermore,we show the molecular mechanisms that govern SPARC-induced neuronal markers and discuss the potential clinicalimpact of SPARC on medulloblastoma tumor genesis.SPARC expression induced the expression of neuronal

markers in medulloblastoma in vitro as shown by immuno-blotting and immunocytochemical analysis. Moreover, Ca2þ

response by high Kþ proved functionally mature neuronactivity in neuronal cells, differentiated from Ad-DsRed-SP–infected medulloblastoma cells. We have also shown that theneuronal differentiation ability of Ad-DsRed-SP depends onSTAT3 regulation. The function of STAT3 in differentiationhas been investigated extensively. Extracellular stimulus and

the cellular context activate the phosphorylation of STAT3.Increasing evidence indicates that STAT3 can maintain thepropagation and pluripotency of embryonic stem cells (32).Suppression of STAT3 directly induced neurogenesis in neuralstem cells (33). STAT3 activation has also been reported to besufficient to maintain the undifferentiated state of mouseembryonic stem cells (28, 34).

Our results further emphasize the potential clinical impor-tance of Notch signaling in medulloblastoma. A recent seriesof experiments showed an oncogenic role for Notch signalingin medulloblastoma and themaintenance of medulloblastomastem cells. The Notch pathway also plays an important role ininhibition of differentiation in many systems, including the

Mock Ad-DsRed Ad-DsRed-SP

H&

ES

PA

RC

No

tch

1p

STA

T3(

Y-7

05)

NF

Neu

NN

esti

nM

AP

-2Figure 7. SPARC overexpression induces expression of neuronal markersin vivo. Orthotopic medulloblastoma tumor sections from mice thatreceived mock, Ad-DsRed, and Ad-DsRed-SP treatments were analyzed.Tissue sections were stained with H&E. Immunohistochemical analysiswas conducted with antibodies specific to SPARC, Notch1, pSTAT3, NF,NeuN, nestin, and MAP-2.

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4917

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

hematopoietic system (35). Notch blockade suppressedexpression of the pathway target HES1 and caused cell-cycleexit, apoptosis, and differentiation in medulloblastoma celllines (6). In addition, expression of the Notch pathway targetgene HES1 in medulloblastomas was associated with signifi-cantly shorter patient survival (36). It was shown that STAT3 isactivated in the presence of active Notch, as well as the Notcheffectors HES1 and HES5 (28). Furthermore, suppression ofendogenous HES1 expression reduces growth factor inductionof STAT3 phosphorylation (28). We therefore tested the effectof SPARC on Notch/HES1-mediated regulation of STAT3phosphorylation. HES1 overexpression induced STAT3 phos-phorylation and suppressed expression of neuronal markers,suggesting that HES1/STAT3 axis plays a role in SPARC-induced neuronal markers. As consistent with our immuno-blot analysis, morphologic and immunocytochemical analysisalso suggest that HES1 overexpression suppressed neuron-likemorphologic changes and neuronal marker expression. Ourstudy clearly shows that HES1 overexpression induced STAT3phosphorylation of SPARC-overexpressed cells.

We next investigated how SPARC modulates IL-6 signalingin medulloblastoma cells. SPARC had a negative regulatoryrole on levels of IL-6sR, an IL-6 agonist implicated in IL-6trans-signaling (37). IL-6 is a multifunctional cytokine that hasalso been implicated in tumorigenesis (38). Cytokines of theIL-6 family were suggested to block neuronal markers expres-sion of cerebral cortical precursor (39). Our study shows thatSPARC expression decreased IL-6 expression. In addition, wealso show that overexpressing IL-6 blocked SPARC-mediatedinhibition of Notch1 expression and neuronal markers expres-sion. Conversely, we also show that SP-siRNA induced IL-6 andNotch expression. Furthermore, we show that blocking IL-6signaling in SPARC-suppressed cells induced Notch1 expres-sion and neuronal differentiation. These findings suggest thatSPARC negatively regulates IL-6 signaling leading to thesuppression of Notch1 signaling, resulting in neuronal differ-entiation of medulloblastoma cells.

Immunohistochemistry showed that cells expressing SPARCexpress high levels of neural markers in the tumor sections ofmice treated with Ad-DsRed-SP. This change in histologicappearance was also associated with a change in the size ofxenograft tumors that formed in the immunodeficient mice(18), suggesting that SPARC enhanced the expression of neu-ronalmarkers inmedulloblastomacells. Ourobservationcanbeplaced within the larger context of recent progress in cancertreatment involving differentiation. In many cell lines andprimary cultures derived from hematologic malignancies, themalignant phenotype can be abrogated by inducing differentia-tion (40). Cyclopamine, a plant-derived teratogen that targetsthe SHH pathway, inhibits SHH-dependent gene expression in

medulloblastoma in vitro and is able to cause cell-cycle arrestconsistent with the initiation of neuronal differentiation andloss of neuronal stem cell–like character (41).

In summary, we have previously shown that SPARCexpression causes tumor growth inhibition (41). Further-more, we show that SPARC induced neuronal differentiation,which could render these tumors to be more susceptible tochemo- and radiotherapy. Previous studies show that SPARCenhances apoptosis in therapy-refractory MIP101 coloncancer cells exposed to chemotherapy by activating theextrinsic pathway of apoptosis while further enhancingthe effect of chemotherapy through the intrinsic pathway(42). The effect of radiotherapy in combination with SPARCis the focus of the ongoing research in our laboratory. Thereare some medically relevant implications from our in vitroand in vivo data. First, SPARC expression can play a role inthe clinical outcome of medulloblastoma patients byincreasing the number of nonproliferative cells that havedifferentiated into neurons. Second, pharmacologic inter-ventions aimed at the mechanisms of medulloblastomadifferentiation outlined in this study might be therapeuti-cally relevant. Third, our results raise the question whetherSPARC also plays a role in the differentiation of other tumortypes such as colorectal cancer or ovarian cancer in whichSPARC has been identified as a potential therapeutic target.Finally, neuronal differentiation of these cells can sensitizetumors for therapy.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgements

The authors thank Shellee Abraham for technical assistance and DianaMeister and Sushma Jasti for manuscript review. The authors also thank Prof. J.E. Darnell, Jr., for providing constitutively active STAT3 (Addgene plasmid:#8722); Dr. Darell D. Bigner, Duke University Medical Center, for providing D425,H2405, and H2411 medulloblastoma cells; Dr. Ali-Osman, Duke UniversityMedical Center, for providing UW228 cells; and Dr. Dan Fults, ComprehensiveCare for Brain and Spine Disorders, Salt Lake City, UT, for providing paraffin-embedded human medulloblastoma tumor sections.

Grant Support

This research was supported by National Cancer Institute grant CA132853(to S.S. Lakka). Contents of this article are solely the responsibility of the authorsand do not necessarily represent the official views of NIH.

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

Received September 15, 2010; revised May 6, 2011; accepted May 17, 2011;published OnlineFirst May 25, 2011.

References1. Crawford JR, MacDonald TJ, Packer RJ. Medulloblastoma in child-

hood: new biological advances. Lancet Neurol 2007;6:1073–85.2. Zeltzer PM, Boyett JM, Finlay JL, Albright AL, Rorke LB, Milstein JM,

et al. Metastasis stage, adjuvant treatment, and residual tumor areprognostic factors for medulloblastoma in children: conclusions from

the Children's Cancer Group 921 randomized phase III study. J ClinOncol 1999;17:832–45.

3. Ehrbrecht A, Muller U, Wolter M, Hoischen A, Koch A, RadlwimmerB, et al. Comprehensive genomic analysis of desmoplastic medul-loblastomas: identification of novel amplified genes and separate

Bhoopathi et al.

Cancer Res; 71(14) July 15, 2011 Cancer Research4918

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

evaluation of the different histological components. J Pathol2006;208:554–63.

4. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A,et al. The 2007 WHO classification of tumours of the central nervoussystem. Acta Neuropathol 2007;114:97–109.

5. Artavanis-Tsakonas K, Riley EM. Innate immune response to malaria:rapid induction of IFN-gamma from human NK cells by live Plasmo-dium falciparum-infected erythrocytes. J Immunol 2002;169:2956–63.

6. Fan X, Matsui W, Khaki L, Stearns D, Chun J, Li YM, et al. Notchpathway inhibition depletes stem-like cells and blocks engraftment inembryonal brain tumors. Cancer Res 2006;66:7445–52.

7. Yan Q, Sage EH. SPARC, a matricellular glycoprotein with importantbiological functions. J Histochem Cytochem 1999;47:1495–506.

8. HuynhMH, Sage EH, RinguetteM. A calcium-bindingmotif in SPARC/osteonectin inhibits chordomesoderm cell migration during Xenopuslaevis gastrulation: evidence of counter-adhesive activity in vivo. DevGrowth Differ 1999;41:407–18.

9. Chlenski A, Liu S, Guerrero LJ, Yang Q, Tian Y, Salwen HR, et al.SPARC expression is associated with impaired tumor growth, inhib-ited angiogenesis and changes in the extracellular matrix. Int J Cancer2006;118:310–6.

10. Brown AL, Ringuette MJ, Prestwich GD, Bagli DJ, Woodhouse KA.Effects of hyaluronan and SPARC on fibroproliferative eventsassessed in an in vitro bladder acellular matrix model. Biomaterials2006;27:3825–35.

11. Lane TF, Sage EH. The biology of SPARC, a protein that modulatescell-matrix interactions. FASEB J 1994;8:163–73.

12. Martinek N, Shahab J, Sodek J, Ringuette M. Is SPARC an evolu-tionarily conserved collagen chaperone?J Dent Res 2007;86:296–305.

13. Ford R, Wang G, Jannati P, Adler D, Racanelli P, Higgins PJ, et al.Modulation of SPARC expression during butyrate-induced terminaldifferentiation of cultured human keratinocytes: regulation via a TGF-beta-dependent pathway. Exp Cell Res 1993;206:261–75.

14. Hrabchak C, Ringuette M, Woodhouse K. Recombinant mouseSPARC promotes parietal endoderm differentiation and cardiomyo-genesis in embryoid bodies. Biochem Cell Biol 2008;86:487–99.

15. Cho WJ, Kim EJ, Lee SJ, Kim HD, Shin HJ, Lim WK. Involvement ofSPARC in in vitro differentiation of skeletal myoblasts. BiochemBiophys Res Commun 2000;271:630–4.

16. Martinek N, Shahab J, Saathoff M, Ringuette M. Haemocyte-derivedSPARC is required for collagen-IV-dependent stability of basal lami-nae in Drosophila embryos. J Cell Sci 2008;121:1671–80.

17. Lawinger P, Venugopal R, Guo ZS, Immaneni A, Sengupta D, Lu W,et al. The neuronal repressor REST/NRSF is an essential regulator inmedulloblastoma cells. Nat Med 2000;6:826–31.

18. Bhoopathi P, Chetty C, Gujrati M, Dinh DH, Rao JS, Lakka SS.Cathepsin B facilitates autophagy mediated apoptosis in SPARCoverexpressed primitive neuroectodermal tumor cells. Cell DeathDiffer 2010;17:1529–39.

19. Bigner SH, Friedman HS, Vogelstein B, Oakes WJ, Bigner DD. Ampli-fication of the c-myc gene in human medulloblastoma cell lines andxenografts. Cancer Res 1990;50:2347–50.

20. Ali-Osman F, Stein DE, Renwick A. Glutathione content and glu-tathione-S-transferase expression in 1,3-bis(2-chloroethyl)-1-nitro-sourea-resistant human malignant astrocytoma cell lines. CancerRes 1990;50:6976–80.

21. Mohan PM, Barve M, Chatterjee A, Hosur RV. pH driven conforma-tional dynamics and dimer-to-monomer transition in DLC8. ProteinSci 2006;15:335–42.

22. Bhoopathi P, Chetty C, Kunigal S, Vanamala SK, Rao JS, Lakka SS.Blockade of tumor growth due tomatrix metalloproteinase-9 inhibition

is mediated by sequential activation of beta1-integrin, ERK, and NF-kappaB. J Biol Chem 2008;283:1545–52.

23. George J, Dravid SM, Prakash A, Xie J, Peterson J, Jabba SV, et al.Sodium channel activation augments NMDA receptor function andpromotes neurite outgrowth in immature cerebrocortical neurons. JNeurosci 2009;29:3288–301.

24. Chetty C, Bhoopathi P, Lakka SS, Rao JS. MMP-2 siRNA inducedFas/CD95-mediated extrinsic II apoptotic pathway in the A549 lungadenocarcinoma cell line. Oncogene 2007;26:7675–83.

25. Nicoll RA, Malenka RC. Contrasting properties of two forms of long-term potentiation in the hippocampus. Nature 1995;377:115–8.

26. Yu LJ,WuML, Li H, Chen XY,WangQ, Sun Y, et al. Inhibition of STAT3expression and signaling in resveratrol-differentiated medulloblas-toma cells. Neoplasia 2008;10:736–44.

27. Shimojo H, Ohtsuka T, Kageyama R. Oscillations in notch signalingregulate maintenance of neural progenitors. Neuron 2008;58:52–64.

28. Kamakura S, Oishi K, Yoshimatsu T, Nakafuku M, Masuyama N,Gotoh Y. Hes binding to STAT3 mediates crosstalk between Notchand JAK-STAT signalling. Nat Cell Biol 2004;6:547–54.

29. WangM,Wu L, Wang L, Xin X. Down-regulation of Notch1 by gamma-secretase inhibition contributes to cell growth inhibition and apoptosisin ovarian cancer cells A2780. Biochem Biophys Res Commun2010;393:144–9.

30. Said NA, Najwer I, SochaMJ, Fulton DJ, Mok SC,Motamed K. SPARCinhibits LPA-mediatedmesothelial-ovarian cancer cell crosstalk. Neo-plasia 2007;9:23–35.

31. Sansone P, Storci G, Tavolari S, Guarnieri T, Giovannini C, Taffurelli M,et al. IL-6 triggers malignant features in mammospheres from humanductal breast carcinoma and normal mammary gland. J Clin Invest2007;117:3988–4002.

32. Matsuda T, Nakamura T, Nakao K, Arai T, Katsuki M, Heike T, et al.STAT3 activation is sufficient to maintain an undifferentiated state ofmouse embryonic stem cells. EMBO J 1999;18:4261–9.

33. Gu F, Hata R, Ma YJ, Tanaka J, Mitsuda N, Kumon Y, et al. Suppres-sion of Stat3 promotes neurogenesis in cultured neural stem cells. JNeurosci Res 2005;81:163–71.

34. Yoshimatsu T, Kawaguchi D, Oishi K, Takeda K, Akira S, MasuyamaN, et al. Non-cell-autonomous action of STAT3 in maintenance ofneural precursor cells in the mouse neocortex. Development2006;133:2553–63.

35. Maillard I, Fang T, Pear WS. Regulation of lymphoid development,differentiation, and function by the Notch pathway. Annu Rev Immunol2005;23:945–74.

36. Fan X, Mikolaenko I, Elhassan I, Ni X, Wang Y, Ball D, et al. Notch1 andnotch2 have opposite effects on embryonal brain tumor growth.Cancer Res 2004;64:7787–93.

37. Kallen KJ. The role of transsignalling via the agonistic soluble IL-6receptor in human diseases. Biochim Biophys Acta 2002;1592:323–43.

38. Hodge DR, Hurt EM, Farrar WL. The role of IL-6 and STAT3 ininflammation and cancer. Eur J Cancer 2005;41:2502–12.

39. Bonni A, Sun Y, Nadal-Vicens M, Bhatt A, Frank DA, Rozovsky I, et al.Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science 1997;278:477–83.

40. Ferrari AC, Waxman S. Differentiation agents in cancer therapy.Cancer Chemother Biol Response Modif 1994;15:337–66.

41. Romer J, Curran T. Targeting medulloblastoma: small-molecule inhi-bitors of the Sonic Hedgehog pathway as potential cancer therapeu-tics. Cancer Res 2005;65:4975–8.

42. Tai IT, Dai M, Owen DA, Chen LB. Genome-wide expression analysisof therapy-resistant tumors reveals SPARC as a novel target forcancer therapy. J Clin Invest 2005;115:1492–502.

SPARC Induces Neuronal Differentiation

www.aacrjournals.org Cancer Res; 71(14) July 15, 2011 4919

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395

2011;71:4908-4919. Published OnlineFirst May 25, 2011.Cancer Res   Praveen Bhoopathi, Chandramu Chetty, Ranadheer Dontula, et al.   Cells via the Notch1/STAT3 PathwaySPARC Stimulates Neuronal Differentiation of Medulloblastoma

  Updated version

  10.1158/0008-5472.CAN-10-3395doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2011/05/25/0008-5472.CAN-10-3395.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/71/14/4908.full#ref-list-1

This article cites 42 articles, 13 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/71/14/4908.full#related-urls

This article has been cited by 3 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

  .pubs@aacr.org

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/71/14/4908To request permission to re-use all or part of this article, use this link

on March 12, 2019. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 25, 2011; DOI: 10.1158/0008-5472.CAN-10-3395