selective release of a cyclopamine glucuronide prodrug toward stem … · cyclopamine has inhibited...

Small Molecule Therapeutics

Selective Release of a Cyclopamine Glucuronide Prodrugtoward Stem-like Cancer Cell Inhibition in Glioblastoma

Anaïs Balbous1,2,3, Brigitte Renoux4, Ulrich Cortes1,2,3, Serge Milin5, Karline Guilloteau1,2,3,Thibaut Legigan4, Pierre Rivet3, Odile Boissonnade6, S�ebastien Martin3, Caroline Tripiana6,Michel Wager7, Ren�e Jean Bensadoun6, S�ebastien Papot4, and Lucie Karayan-Tapon1,2,3

AbstractRecent data suggest that inhibition of the Hedgehog pathway could be a therapeutic target for

glioblastoma. Alkaloid cyclopamine inhibits Hedgehog signaling, depleting stem-like cancer cells derived

from glioblastoma. However, this compound is toxic for somatic stem cells, preventing its use for clinical

applications. In this study, we tested a derivatization product of cyclopamine in the form of cyclopamine

glucuronide prodrug (CGP-2). This compound was used in vitro and in vivo toward glioblastoma-initiating

cells (GIC). Results obtained in vitro indicate that CGP-2 is active only in the presence of b-glucuronidase,an enzyme detected in high levels in necrotic areas of glioblastomas. CGP-2 decreased proliferation

and inhibited the self-renewal of all GIC lines tested. Hedgehog pathway blockade by 10 mmol/L of CGP-2

induced a 99% inhibition of clonogenicity on GICs, similar to cyclopamine treatment. Combination of

CGP-2 with radiation decreased clonogenic survival in all GIC lines compared with CGP-2 alone. In a

subcutaneous glioblastoma xenograft model, a two-week CGP-2 treatment prevented tumor growth

with 75% inhibition at 8 weeks, and this inhibition was still significant after 14 weeks. Unlike cyclopamine,

CGP-2 had no detectable toxic effects in intestinal crypts. Our study suggests that inhibition of the

Hedgehog pathway with CGP-2 is more effective than conventional temozolomide adjuvant, with much

lower concentrations, and seems to be an effective therapeutic strategy for targeting GICs.Mol Cancer Ther;

13(9); 2159–69. �2014 AACR.

IntroductionGlioblastoma multiforme (GBM; World Health Orga-

nization grade IV glioma) is the most frequent andaggressive primary brain tumor in adults. Standardtreatment consists of surgical resection of tumor, fol-lowed by adjuvant radiation therapy and chemothera-py using temozolomide, an oral alkylating agent. How-ever, GBM remains one of the most fatal and leastsuccessfully treated solid tumors with a median sur-vival of 14.6 months with current treatment (1). Studies

have suggested that the progression of GBM is associ-ated with presence in the tumor of a small subpopula-tion of cells with stem cell characteristics called "glio-blastoma-initiating cells" (GIC; refs. 2, 3). These reportshave indicated that tumorigenic cells in GICs wererestricted to the CD133þ population. However, recentstudies have revealed that CD133-negative cells isolat-ed from GBM could also be tumorigenic (4–6) and mayin fact represent interconvertible phenotypic states ofthe same cell population giving rise to a heterogeneousGIC population (4). These cells are highly resistant toconventional chemotherapeutic drugs, including temo-zolomide (7, 8) and mediate tumor recurrence follow-ing radiation therapy (9). Therefore, complete eradica-tion of GICs might be a prerequisite for successfultherapeutic strategies.

Signaling pathways involved in the proliferativeand self-renewal capacities of stem cells are a potentialtarget in attempts to reduce relapse and improve patientsurvival. Transient activation of the Hedgehog (HH)and Wnt pathways is known to promote stem cellself-renewal in normal tissues, whereas continuousactivation is associated with initiation of many typesof cancer such as skin (10), lung (11), prostate (12),gastrointestinal tract (13), breast (14), and brain can-cer, including GBM (10, 15). Sandberg and colleagues

1INSERMU935, Mod�eles de cellules souches malignes et th�erapeutiques,Poitiers, France. 2Universit�e de Poitiers, Poitiers, France. 3CHUde Poitiers,Laboratoire de Canc�erologie Biologique, Poitiers, France. 4Universit�e dePoitiers, UMR-CNRS 7285, Institut deChimie desMilieux et desMat�eriaux,Groupe «Syst�emes Mol�eculaires Programm�es, Poitiers, France. 5CHU dePoitiers, Service d'Anatomo-cytopathologie, Poitiers, France. 6CHU dePoitiers, Service d'Oncologie Radiotherapique, Poitiers, France. 7CHU dePoitiers, Service de Neurochirurgie, Poitiers, France.

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Corresponding Author: Lucie Karayan-Tapon, Laboratoire de Can-c�erologie Biologique, Centre Hospitalier Universitaire de Poitiers, Poi-tiers F-86021, France. Phone: 33549444988; Fax: 549444095; E-mail:[email protected]

doi: 10.1158/1535-7163.MCT-13-1038

�2014 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 2159

on May 27, 2021. © 2014 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst July 22, 2014; DOI: 10.1158/1535-7163.MCT-13-1038

http://mct.aacrjournals.org/

recently identified alterations of Wnt and HH signalingpathways as important events conferring self-renewalpotential and tumorigenic properties to GICs (16). Inaddition, it has been shown that the HH pathway isimplicated in glioma resistance to alkylating agents andthe maintenance of GICs (17–19).

Cyclopamine is a natural alkaloid from Veratrum cali-fornicum (California corn-lily) that is known to inhibit theHHpathway by direct binding on the heptahelical bundleof Smo (20–22). This compound exhibits antiproliferativeactivity on glioma cell lines (17). Recent data have dem-onstrated that blockage of the HH pathway by cyclopa-mine reduces proliferation, neurosphere formation, self-renewal, and tumorigenicity of GICs with or withoutradiation (18, 19). Furthermore, pretreatment ofGICswithcyclopamine has inhibited tumor formation and growthin athymicmice (18). In addition, cyclopaminewas recent-ly evaluated for its antitumor activity in preclinical andclinical trials (23).

Although cyclopamine seems to be a promising che-motherapeutic agent, this drug could be highly toxicfor somatic stem cells in normal tissues like brain, whichare also HH dependent. In an attempt to circumventthis drawback, Hamon and colleagues have recentlydesigned glucuronide prodrugs programmed to delivercyclopamine selectively in the tumor microenvironment(24–26). In fact, several studies have shown that glucu-ronide prodrugs can be activated by b-glucuronidasepresent in high concentrations in necrotic areas of mostsolid tumors, including glioblastomas (27). b-Glucuron-idase is secreted extracellularly in tumors by inflamma-tory cells such as neutrophils and macrophages, while inhealthy tissues, its activity is confined to lysosomes(28, 29). This enzymatic specificity of the tumor micro-environment allows glucuronide prodrugs to discrimi-nate between malignant and normal tissues. As expect-ed, the derivatization of cyclopamine in the form ofcyclopamine glucuronide prodrugs (CGP) significantlyreduced its toxicity. When incubated with b-glucuron-idase, CGPs led to release of the free drug, therebyrestoring its antiproliferative activity against U87-MGglioblastoma cell line. In light of these promising pre-liminary results, it seemed worthwhile to pursue eval-uation of this targeting strategy as a potential alternativemeans of glioma treatment. Indeed, this therapeuticapproach can be used in a systemic way since it hasbeen recently reported that the blood–brain barrier(BBB) is disturbed in GBM (30, 31). Newly formed bloodvessels in GBM display abnormal structure and loss ofexpression of tight junctions and claudin-3 causingstructural and functional alterations of the BBB (32).These alterations allow chemotherapeutics to reach thebulk of the tumor. Furthermore, the BBB permeabilitycan also be increased using ultrasound technologyallowing transient opening and drug delivery withoutneural cells damage (33). Within this framework, the aimof the present study is to appraise the effect of the CGP-2on glioma-initiating cells in vitro and in vivo.

Materials and MethodsCyclopamine glucuronide prodrug production

The CGP-2 (Supplementary Fig. S1) was selected asthe best candidate for this study due to its improvedaqueous solubility compared with its previous ana-logue CGP-1 described by Hamon and colleagues(24) CGP-2 was prepared following the previouslydescribed synthetic strategy (25). The compound wasthen purified using preparative high-performance liq-uid chromatography before biologic evaluation (purity> 95%). In our experiments, CGP-2 was dissolved inDMSO.

GIC cell lines, H9-NSC, and cell cultureTumor samples were obtained within 30 minutes

after surgical resection from 4 adult patients with glio-blastoma (GIC-2, GIC-3, GIC-6, and GIC-12). The meth-odology for isolation and characterization of these cellshas been described in details in our earlier reports (34–36). All GIC lines were assessed for self-renewal, dif-ferentiation, and in vitro clonogenicity by limiting dilu-tion assays. In addition, tumorigenicity and stemnessproperties of GBM-derived initiating cells were evalu-ated by xenograft experiments in nude mice. Cellsderived from all four tumors were cultured as prolif-erative nonadherent spheres in neurobasal medium(NBE) supplemented with 20 ng/mL of basic fibroblastgrowth factor (bFGF; Life Technologies), 20 ng/mL ofEGF (Life Technologies), and culture supplements N2(100X, Life Technologies) and B27 (50X, Life Technolo-gies). The molecular characteristics, including MGMTpromoter methylation, EGFR copy number, IDH1,IDH2, EGFR-variant III, p53, PTEN status as well asLOH at loci 1p36, 19q13, 9p21, and 10q23, of GICs andtheir corresponding tumors are notified in Supplemen-tary Table S1. Human neural stem cells (H9-NSCs) arederived from NIH-approved H9 (WA09) human embry-onic stem cells (Life Technologies). Cells were culturedfollowing the manufacturer’s instructions.

Reverse transcription quantitative PCRTotal RNA was extracted from tumors and GICs

by RNAqueous-4PCR Kit (Life Technologies) accordingto the manufacturer’s instructions. First-strand cDNAsynthesis was carried out on 1-mg samples of totalRNA using High Capacity cDNA Reverse TranscriptionKit with RNase Inhibitor (Life Technologies). SHH,PTCH1, SMO, GLI1, and GLI3mRNA expressionswere determined using TaqMan gene expression assaysHs00179843_m1, Hs00181117_m1, Hs00170665_m1,Hs01110768_m1, and Hs00609233_m1 (Life Technolo-gies), respectively. qRT-PCR was carried out in dupli-cate and performed using the TaqMan master mix (LifeTechnologies) on a 7900HT Fast system (Life Technol-ogies). SHH, PTCH1, SMO, GLI1, and GLI3mRNAexpressions were normalized to GAPDH mRNA levelsand determined by the 2�DDCt method.

Balbous et al.

Mol Cancer Ther; 13(9) September 2014 Molecular Cancer Therapeutics2160




Cell viability assaysThe CellTiter96 Aqueous Non-Radioactive Cell Prolif-

eration Assay (Promega) was used to determine the effectof cyclopamine (Biovalley), temozolomide (Interchim),and CGP-2 on cell viability. Cells were plated in 96-wellplates at a density of 5 � 104 cells per well in 90-mLmedium. After 24 hours of incubation, 10 mL of CGP-2� b-glucuronidase (40 IU/mL, Sigma-Aldrich), cyclopa-mine, or temozolomide was added to each well. Controlcells were treated with DMSO or b-glucuronidase. Cellviability was determined after 5 days of treatment. Thequantification of viable cells was performed at 492 nmwith a microplate reader (Dynex Technologies). The IC50

value was calculated as the drug concentration requiredto inhibit cell viability by 50% compared with control(DMSO or b-glucuronidase).

Cell proliferation assaysCellswere plated in 96-well plates at a density of 5� 104

cells per well in 90-mL medium. After 24 hours of incu-bation, 10 mL of CGP-2 (at IC50 concentration) þ b-glucu-ronidase (40 IU/mL) were added to each well. Controlcells were treated with DMSO þ b-glucuronidase. Thequantification of viable cells was performed at 492 nmwith a microplate reader at indicated times.

Clonogenic assays in methycelluloseGIC cells (4� 104 cells) were plated in 30-mmPetri dish

(Greiner Bio One) in 1 mL of complete methylcellulosemedium S1 (Stem Cell Technologies) and supplementedwith B27 (50X), N2 supplements (100X), bFGF, and EGF(0, 1 mg/mL). GIC-2, -3, -6, and -12 had consistently lowplating efficiencies (7.4%, 2.8%, 8.1%, and 7.8%, respec-tively) and therefore required high seeding concentra-tions to provide sufficient number of cells to be scoredafter treatment. Cells were treated with CGP-2 (5 and10 mmol/Lþ 40 IU/mL of b-glucuronidase), cyclopamine(5 and 10 mmol/L), or temozolomide (200 and 800mmol/L). After 21 days of incubation, colonies consistingof more than 50 cells were counted under an invertedmicroscope. The experimentwas repeated three times. Forradiation combination, the cells were plated in methyl-cellulose, treated with CGP-2 (IC50 concentration þ 40IU/mL of b-glucuronidase) or temozolomide (IC50 con-centrations), and irradiated using an Elekta SynergyBeam Modulator (electron energies 18 MeV) at a dose of2 Gy and colonies were counted after 21 days.

Limiting dilutions and neurosphere-initiating cellassaysTo determine the effect of CGP-2 and temozolomide on

the frequency of neurosphere-initiating cells (NS-IC), weperformed limiting dilution assays using single cell popu-lations immediately after dissociation of neurospheres.Final cell dilutions ranged from 1 cell per well to 100 cellsper well (IC50 concentrations). After a 21-day incubation,the fraction of wells not containing neurosphere wasdetermined at each plating density. The fraction of neg-

ative wells was plotted against the number of cells perwell. The number of cells required to form one neuro-sphere, which reflects the frequency of cancer-initiatingcells in the entire population, was determined asdescribed previously (34, 35).

Annexin V and flow cytometryGICs were seeded 24 hours before treatment with 500

mmol/L temozolomide or 30 mmol/L CGP-2 with b-glucu-ronidase (40 IU/mL). After 5 days, cells were stained withAnnexin V and 7-aminoactinomycin D (7-AAD) using aFITC Annexin V apoptosis detection kit (BD Biosciences)following the manufacturer’s instructions. The cells werewashed twice with PBS and stained with FITC-conjugatedAnnexin V (5 mL) and 7-AAD (7 mL, BD Biosciences) in100 mL of binding buffer [10 mmol/L Hepes/NaOH(pH7.4), 140 mmol/L NaCl, 2.5 mmol/L CaCl2] for 15minutes at room temperature in the dark. Apoptosis wasmeasured immediately by flow cytometry on a FACSCanto II (BD Biosciences). Data analysis was performedusing FACS Diva software (BD Biosciences). A total of10,000 events were analyzed in one typical experiment.

Mouse xenograftsFemale athymic Swiss nude mice were used at 9

weeks of age (Crl:NU-Foxn1nu, Charles River). All ofthe mice were maintained in a conventional-specificpathogen-free facility and received standard laboratoryrodent food and water ad libitum. All experiments wereconducted in accordance with institutional guidelinesfor animal use and care. GIC-2 cells (1 � 106) werecultured as previously described (see cell culture sec-tion). The cells were dissociated, suspended in NBE,and injected subcutaneously with a 26-gauge needleinto the left flank of nude mice. Tumors were observedat injection site (>4 mm) after 130 days in 8 of 18 miceand collected when their size reached 1000 mm3. Theywere then removed, finely chopped, and dissociatedinto single cell suspension. Cells were grafted subcuta-neously into the left flank of female nude mice at aconcentration of 3 � 105 cells. Mice were randomlyallocated to the treatment groups when tumor volumereached 42 mm3. They were anesthetized with isoflur-ane (1.5%–3%) and a drop of ophthalmic anesthetic(0.5% proparacaine hydrochloride ophthalmic solution,Sigma-Aldrich) was placed on the eye that wouldreceive the injection. Cyclopamine (4 mg/kg), CGP-2(50 mg/kg), or vehicle (20% DMSO) was injected in theretro-orbital sinus with a 0.5-inch insulin needle andsyringe once every 3 days (5 injections).

To determine tumor growth, tumor size was estimatedwith calipers every week. Tumor volume (V) was calcu-lated asV¼ (a� b2)/2,where a is the long tumor axis and bis the short tumor axis. Mice were sacrificedwhen tumorsreached 17 mm as the longest axis (�2,000 mm3) or onappearance of the first clinical signs (weight loss). Sub-cutaneous tumorswere removed, fixed in 4% paraformal-dehyde, embedded in paraffin, cut in 3 mm-thick sections,

Cyclopamine Glucuronide Prodrug toward Glioma Stem Cells

www.aacrjournals.org Mol Cancer Ther; 13(9) September 2014 2161




and stained with hematoxylin and eosin (H&E). IHCstaining to examine GFAP expression in subcutaneoustumors and patient was performed using GFAP mousemonoclonal antibody (1/200; Dako).

ResultsHeterogeneous expression of the HH pathwaycomponents in glioblastoma-initiating cells

The HH pathway was shown to be active in a subsetof primary GBM and established GBM cell lines (18).To evaluate HH pathway expression in GICs, we exam-ined mRNA expression of pathway receptors SMOand PTCH1, ligand SHH and targets GLI1 and GLI3 byreal-time RT-PCR in four GICs. Expression of thesegenes was then compared with human H9-NSCs andto the corresponding tumors. The HH pathway recep-tors SMO and PTCH1 mRNA were detected at variouslevels, in line with previous results from Bar and col-leagues 2007(18), with a lower expression level ofPTCH1 as compared with SMO. SMO and PTCH1 weremost highly expressed in GIC-3 and GIC-6. Analysis ofGLI1 and GLI3 targets revealed different expressionprofiles with higher levels of GLI1 transcript in GIC-6and GIC-12. Very low levels of SHH mRNA expression

were detected in our GICs. A lack of SHH mRNAexpression had previously been observed in adherentcell lines as well as in high-grade gliomas (18). Theseobservations corroborate previous results obtained inGBM-derived neurospheres (18, 19) and suggest that theHH pathway is present in these cells. SMO, PTCH1, andGLI1 were similarly expressed in GICs and H9-NSCs,whereas GLI3 and SHH expressions were higher in H9-NSCs (Fig. 1A). In the corresponding tumors, the HHpathway receptor SMO and targets GLI1 and GLI3expressions were comparable with GICs. Interestingly,PTCH1 mRNA level was 2- to 3-fold higher in tumorsthan the corresponding GIC. In these tumors, SHHmRNA expression was similar to that of GICs exceptfor GBM-12 showing increased expression (Fig. 1B).

Cyclopamine glucuronide prodrug inhibits viabilityof GICs

We initially examined the effect of cyclopamine onGICs viability. As expected, cyclopamine treatmentresulted in reduced viability of GICs. The viability ofthese cells decreased after 5 days of treatment with IC50

values of 24, 33, 29, and 24 mmol/L for GIC-2, -3, -6, and-12, respectively (Table 1). We then evaluated the effectof CGP-2 on GIC viability. The cells were treated for

Figure 1. HH pathway components are heterogeneously expressed in GICs and in corresponding GBM tumors. A, SMO, PTCH1, GLI1, GLI3, and SHHmRNA were detected in GICs and H9-NSC by qRT-PCR. B, the SHH pathway components were detected in corresponding GBM tumors by qRT-PCR. GADPH was used as an internal control. Values are expressed as mean � SD.

Balbous et al.





5 days either with the prodrug alone or in combinationwith b-glucuronidase, and MTS assays were perform-ed. CGP-2 had no effect on GIC viability when usedalone. In contrast, when prodrug was combined withb-glucuronidase, the viability of GICs markedlydecreased (P < 0.001), with IC50 values between 25 and29 mmol/L (Fig. 2A). We further examined the effect ofCGP-2 (IC50) in combination with b-glucuronidase onGICs proliferation. Analysis of growth curves showed areduction of proliferation in all GICs treated with CGP-2(Fig. 2B; P < 0.001). Nevertheless, at IC50 concentration(30 mmol/L), the prodrug had no effect on H9-NSCs inthe absence of b-glucuronidase (Supplementary Fig. S2).To analyze themechanisms of cell death, we carried out

Annexin staining of GICs after treatment with 30 mmol/Lof CGP-2þ b-glucuronidase. After 5 days of treatment, allGICs underwent apoptosis at various levels. As shownin Fig. 2C, in the absence of CGP-2, similar basal levels ofapoptosis were present in GIC-2, -3, and -12 except forGIC-6. After 5 days of treatment with CGP-2, the percent-age of apoptotic cells had risen significantly from 19% to69%, 20% to 77%, 18% to 40%, and 50% to 70%, respec-tively (P < 0.05). These results are in accordance withprevious findings indicating increased apoptosis follow-ing cyclopamine treatment (19).We carried out RT-PCR experiments after 5 days of

treatment with CGP-2 on two GIC lines (GIC-6 and GIC-12) and results showed, respectively, a 3- and 7-folddecrease in nestin expression after treatment (Supplemen-tary Fig. S3). Analysis of CD44 and Sox2 mRNA expres-sion revealed only slight variations after treatment (datanot show); in contrast, we observed noticeable changes inGFAP expression in GIC-6 although mRNA levelsremained very weak (data not show).

CGP-2 inhibits clonogenicity and neurosphereinitiationTo study the ability of cyclopamine and CGP-2 to block

colony formation, cells were seeded in methylcelluloseafter dissociation and treated with increasing concentra-tions of cyclopamine orCGP-2þ b-glucuronidase. Colonyformation was assessed by counting after 21 days. Thecolony-forming capacity of GIC cells was inhibited withboth drugs in a dose-dependent manner. A significant

reduction of colony growthwas observed in all four of theGIC lines with 5 mmol/L of cyclopamine, as inhibitionranged from 18% (GIC-2) to 90% (GIC-12; Fig. 2D). Similarobservations were made with 5 mmol/L of CGP-2 þb-glucuronidasewith a 37% (GIC-6) to 89% (GIC-2) reduc-tion of colony growth. Inhibition of colony formationof GIC cells rose as high as 99% after treatment with10 mmol/L of cyclopamine or 10 mmol/L of CGP-2 þb-glucuronidase (Fig. 2D).

To determine the effect of CGP-2 on the frequency of"NS-IC", we performed limiting dilution assays usingsingle-cell populations immediately following neuro-sphere dissociation. The frequency of NS-IC, correspond-ing to the number of cells required to generate at least onetumor sphere per well was 1/17, 1/25, 1/16, and 1/42 forGIC-2, -3, -6, and -12, respectively (Table 2). Treatment ofcellswith IC50 doses ofCGP-2 resulted in adecrease ofNS-IC frequency to 1/321 for GIC-3, 1/1,710 for GIC-6 (P <0.001), 1/735 for GIC-2, and 1/86 for GIC-12 (P < 0.05).

Taken together, these data show that derivatization ofcyclopamine in the formof the glucuronide prodrugCGP-2 turned off its activity. Furthermore, the presence ofb-glucuronidase was necessary to restore cyclopamine’sbiologic properties, which include GIC proliferation inhi-bition, clonogenicity, and neurosphere initiation.

CGP-2 in combination with radiationThe standard Stupp protocol for the treatment of

glioblastoma includes tumor resection followed by che-motherapy with temozolomide and concomitant radia-tion therapy. We first compared the effect of temozo-lomide with CGP-2. We have previously shown that theintended clinical dose of temozolomide (50 mmol/L) hasno effect on GIC proliferation (35). Dose–responsecurves of temozolomide on GIC-2, -3, -6, and -12showed IC50 values of 474, 787, 830, and 864 mmol/L,respectively (Table 1). As observed with CGP-2, temo-zolomide also inhibits colony formation. Colony growthwas significantly reduced in all GIC lines in the pres-ence of 200 mmol/L temozolomide (24%–50%; Fig. 3A)and growth inhibition using 800 mmol/L temozolomideranged from 47% to 83% (P < 0.001). The effect oftemozolomide on NS-IC frequency was limited as fre-quency was 1/735 for GIC-2 and 1/661 for GIC-6 (P <

Table 1. IC50 values after 5 days of cyclopamine, CGP-2, or temozolomide treatments

IC50 values ofcyclopamine

IC50 values ofCGP-2

IC50 values oftemozolomide

GIC-2 24 mmol/L (�1.5) 25 mmol/L (�0.2) 474 mmol/L (�47)GIC-3 33 mmol/L (�1.9) 26 mmol/L (�1.8) 787 mmol/L (�54)GIC-6 29 mmol/L (�1.6) 29 mmol/L (�3.9) 830 mmol/L (�40)GIC-12 24 mmol/L (�3.2) 27 mmol/L (�4.6) 864 mmol/L (�1.7)

NOTE: IC50 values of cyclopamine and CGP-2 were much lower than temozolomide IC50 values. IC50 values were calculated as themean drug concentration required for inhibiting cell proliferation by 50% compared with control (mean � SD).






0.05) when cells were treated with IC50 concentrations.Furthermore, we did not observe any significant reduc-tion in NS-IC frequency in GIC-3 and GIC-12 (1/164 and1/63, respectively; Table 2). Taken together, these dataindicate that CGP-2 treatment (in the presence of b-glu-curonidase) was more effective than temozolomide atmuch lower concentrations as it led to enhanced anti-proliferative activity, inhibition of colony formation inGICs, and reduced NS-IC frequency.

We went on to investigate whether or not CGP-2 incombination with radiation was more potent in reducingGIC survival than the conventional adjuvant temozolo-mide. Following 2 Gy radiation, even though clonogenicsurvival of all GIC lines decreased significantly beyond 21

days (P < 0.001), the population of resistant cells remainedhigher than 70% forGIC-3 andGIC-6, 60% forGIC-12, and50% for GIC-2 (Fig. 3B). Combined temozolomide treat-ment (IC50 concentrations) with radiation reduced thenumber of resistant cells in comparison with radiationalone or temozolomide alone (P < 0.001) with their pop-ulation ranging from 18% to 39%. When combined withradiation, IC50 doses of CGP-2 drastically reduced clono-genic survival of all GIC lines comparedwithCGP-2 alone(Fig. 3B; P < 0.001) or radiation alone as the population ofresistant cells dropped below 10% for GIC-2, -3, -6 and16% for GIC-12. Combination of CGP-2 with radiationwas clearly more potent than the conventional temozo-lomide adjuvant (P < 0.001). Additional tests performed

Figure 2. Effects of CGP-2 and cyclopamine on viability, proliferation, apoptosis, and clonogenicity in GICs. A, cell viability was measured using an MTSassay. CGP-2 alone had no effect on GIC viability. After addition of 40 IU/mL b-glucuronidase cell viability decreased in a dose-dependent manner. Eachset of results was obtained from two independent experiments. Experiments were performed in sextuplicate and expressed as mean � SD (��, P < 0.001).B, cell proliferation was measured using an MTS assay. CGP-2 reduced cell proliferation at IC50 concentration. Each set of results was obtained fromtwo independent experiments. Experiments were performed in sextuplicate and expressed as mean � SD (��, P < 0.001). C, GICs were treated with30 mmol/L CGP-2 and 40 IU/mL b-glucuronidase for 5 days and analyzed by flow cytometry with Annexin V/7-AAD labeling. �, P < 0.05, calculatedwith a two-tailed unpaired Student t test. D, colony formation was measured by clonogenicity assay after 21 days of treatment. Of note, 10 mmol/Lcyclopamine or CGP-2 treatment decreased colony numbers up to 99%. Each set of results was obtained from two independent experiments. Experimentswere performed in triplicate and expressed as mean � SD (��, P < 0.001).

Balbous et al.





on GIC-2 and GIC-6 cells indicate that the combination ofCGP2 with temozolomide or temozolomide and radia-tion reduced drastically the number of resistant cells (P <0.001; Fig. 3B).

In vivo effects of CGP-2To test the antitumoral activity of CGP-2 in vivo, we

first established subcutaneous glioblastoma xenografts

in athymic Swiss nudemice (see Materials andMethods).As shown in Fig. 4B, tumor xenografts that originatedby injectingGICs (ii) shared the samehistologic propertiesas the original tumors (i); H&E section of xenografts frompatients showed pseudopalisading surrounding areas ofnecrosis center (NC). These tumors were characterizedby high cellularity, nuclear atypia, and mitotic activity(iii and iv), which reflected the patient’s original tumor

Table 2. Effects of CGP-2 and temozolomide treatment on NS-IC assay

NS-IC frequencycontrol

NS-IC frequency afterCGP-2 treatment

NS-IC frequency aftertemozolomide treatment

GIC-2 1/17 (�4) 1/735 (�28; aP < 0.05) 1/735 (�46; aP < 0.05)GIC-3 1/25 (�3) 1/321 (�55; bP < 0.001) 1/164 (�6)GIC-6 1/16 (�1) 1/1710 (�725; bP < 0.001) 1/661 (�16; aP < 0.05)GIC-12 1/42 (�5) 1/86 (�2; aP < 0.05) 1/63 (�2)

NOTE: Treatment with CGP-2 and temozolomide resulted in a significant decrease of NS-IC frequency (a, P < 0.05; b, P < 0.001;calculated with a two-tailed unpaired Student t test). Final cell dilutions yielding 37% of negative wells correspond to the dilution atwhich there is one NS-IC per well (mean � SD).

Figure 3. CGP-2 and temozolomidein combination with radiation. A,colony formation wasmeasured byclonogenicity assay after 21 daysfollowing 200 and 800 mmol/Ltemozolomide treatment. Thenumber of colonies formed wasmarkedly reduced bytemozolomide. B, combined CGP-2 with radiation was more effectivethan combined temozolomidetreatment. Each set of results wasobtained from two independentexperiments. Experiments wereperformed in triplicate andexpressed as mean � SD.��, P < 0.001, calculated witha two-tailed unpaired Studentt test.






histology. Patient’s tumors (v) and xenografts (vi) expres-sed the astrocyte-differentiation marker GFAP. Whenthe tumors reached a volume of 42mm3,mice were inject-ed through the retro-orbital venous sinus with either100 mL of vehicle (20% DMSO/80% water), cyclopamine(4 mg/kg), or CGP-2 (50 mg/kg) once every 3 days (5

injections). The derivatization of cyclopamine in the formof the glucuronide prodrug CGP-2 enhanced its watersolubilitywith a reduced toxicity allowing administrationof higher doses compared with the free drug. A majorlimitation of this in vivo study was the high cost ofproducing sufficient amounts of glucuronide prodrug.

Figure4. CGP-2 prevented growth of subcutaneous glioblastoma xenografts without side effects on intestinal crypts. A, 1� 106 GIC-2 cells derived froma patients with GBM (GBM-2) were subcutaneously injected into athymic mice. Cyclopamine (n ¼ 4, 4 mg/kg), CGP-2 (n ¼ 4, 50 mg/kg), or vehiclecontrol DMSO (n¼ 4, 20% v/v) were injected when tumor reached 42 mm3. CGP-2 delayed tumoral development for 14 weeks. Results are expressed asmean � SD. B, H&E section of patient's original glioblastoma GBM-2 (i) and of subcutaneous glioblastoma xenografts. GIC-2 derived from GBMshowed pseudopalisading (ii, black arrowheads) surrounding areas of necrosis center (NC), high cellularity, nuclear atypia, and mitotic activity (iii and iv,white arrowhead). These histologic characteristics reflect the corresponding original patient tumors. Patient tumors (v) and xenografts (vi) were positivefor GFAP staining. C, representative section of intestinal crypts from prodrug-treated mice showing normal histology. One of four cyclopamine-treatedmice showed apoptotic cells (black arrowheads) in small intestinal crypts. Original magnifications: �200 (ii, v, vi, vii) �400 (i, iv, viii).

Balbous et al.





For this reason, only 4 mice were included in each treat-ment arm. As shown in Fig. 4A, treatment with cyclopa-mine or CGP-2 prevented tumor growth with 75% inhi-bition at 8 weeks compared with DMSO-treated mice.However, even though short-term protection wasobserved with both compounds, long-term follow-upobservations indicated a noticeable resumption of tumor-al development in cyclopamine and CGP-2–treated micewith 70% inhibition after 11 weeks but less than 50% after14 weeks. Median tumor volumes for cyclopamine andCGP-2–treated mice were similar to the DMSO controlgroup beyond 20 weeks. Previous reports from van denBrink and colleagues (37, 38) indicate that cyclopaminetreatment disturbed enterocyte maturation in colonic epi-thelium with decreased proliferation in the crypts of thesmall intestine and increased proliferation of the gastricepithelium. To determine whether or not intravenousinjectionof cyclopamine orCGP-2 affected intestinal cryptmorphology, we performed H&E staining of the cryptarea. As shown in Fig. 4C, no morphologic changeoccurred in the small intestine of prodrug-treated mice(vii), but cell apoptosis in the crypts of the small intestinewas observed in 1 of the 4 mice treated with cyclopamine(viii). Quantitative analysis of intestinal toxicity revealedthe presence of heterogenous pathologic areas in a cyclo-pamine-treated mouse containing up to 10% of apoptoticcells. In contrast, in CGP-2–treated mice, apoptotic cellnumber was far below 1%.

DiscussionThe HH pathway is critical to maintenance of stem

cell self-renewal in many tissues. Its aberrant hyperac-tivation is found in different types of cancer, particu-larly in glioblastoma (15, 17). Cyclopamine was the firstHh inhibitor to be identified (20) and its mechanism ofaction is well established (21). A number of preclinicaland clinical studies have been performed using cyclo-pamine and synthetic SMO antagonists (23, 39). Studiesusing disease animal models have reinforced the poten-tial of cyclopamine as an anticancer drug (19, 40–43).Indeed, cyclopamine has the capacity to inhibit pro-liferation of GICs in vitro and to reduce their tumori-genicity in vivo (18, 19). It nonetheless exhibits poorsolubility, while rapid clearance and nonspecific toxic-ity limit its usefulness as a chemotherapeutic agent. Itconsequently appeared worthwhile to develop a newcyclopamine-based drug with both reduced toxicitytoward healthy tissues and an improved pharmacoki-netic profile. Recently, Renoux and colleagues (25) de-signed a new water-soluble glucuronide cyclopamineprodrug (CGP-2) enabling the b-glucuronidase–cata-lyzed release of the drug to overcome its unselectivetoxicity. They demonstrated the validity of this in vitroapproach on a U87-MG glioma cell line. As an extensionof this type of preliminary approach, the aim of ourstudy was to test the effect of CGP-2 on GICs in vitro andon xenografted tumors in vivo.

In all four of the tested GIC lines and in the correspond-ing tumors, the key components required to transduceHH signaling were present and heterogeneously express-ed in accordance with previous reports (18, 19). SHHligand mRNA was weakly expressed in GICs, a findingthat partially explained the higher expression of GLI3compared with GLI1; although transcription of GLI1 issilenced in the absence of SHH ligand, GLI3 can still beexpressed (44). Furthermore, inhibition of GLI3 has beenshown to revert the antiproliferative and proapoptoticeffects of cyclopamine,whichwere therebymediated (19).

As expected, CGP-2 is active only in the presence ofb-glucuronidase and has the capacity to decrease prolif-eration and to prevent the self-renewal and clonogenicityof all GIC lines tested in vitro. In the range of concentra-tions tested, CGP-2 is more active than temozolomide. Inaccordance with previous work by Beier and colleagues(45), the intended clinical doses of temozolomide (50mmol/L in plasma and 10 mmol/L in cerebrospinal fluid;refs. 36, 37) are not sufficient to prevent proliferation. Asregards, the clonogenic properties of GIC lines in vitro, theabove-mentioned effects have been observed only at pro-nouncedly higher concentrations (500 mmol/L).

The preliminary experiments we conducted in a sub-cutaneous xenograft model also indicate that similarCGP-2 antitumoral activity is observed in vivo. Histologicanalysis of xenografted tumors showed that the mousemodel recapitulated most of the typical features de-scribed in GBM. Intravenous administration of cyclopa-mine and CGP-2 significantly prevented tumor growthin mice over 8 weeks, but the effect could not be sus-tained on a long-term basis. Results were neverthelessvery promising because the prodrug was administeredonly over a period of 2 weeks; in the future, it could beworthwhile to try to determine the effects of CGP-2 overthe course of long-term treatment. In addition, unlikecyclopamine, no toxicity signs were noted in mice trea-ted with the prodrug, in accordance with locally con-trolled delivery procedures. Furthermore, in vitro andin vivo results indicate that derivatization of cyclo-pamine in the form of CGP-2 fully abolished its off-targettoxicity but did not alter its efficacy on tumor cells. Witha favorable toxicity profile and promising efficacy, CGP-2 would appear to be a good candidate for testing inclinical applications. In addition, our results providein vitro evidence of the relatively high efficiency ofCGP-2 combined with 2 Gy radiation.

Over recent years, numerous efforts have been devotedto enhancing the efficacy of b-glucuronidase-responsiveprodrugs. For instance, Juan and colleagues (29) demon-strated that antiangiogenic agents can synergize the anti-tumor activity of glucuronide prodrugs by selectivelyenhancing concentration of the activating enzyme in thetumor microenvironment. More recently, the concept ofglucuronidase-responsive albumin-binding prodrugsseemed to represent a promising approach to limiting therapid renal clearance observedwithprevious glucuronideprodrugs (46). These novel approaches could be used in






the near future to increase the therapeutic potential ofCGPs.

To sum up, our experimental results suggest that tar-geting of the Hh pathway could represent a valid strategyand serve as an adjuvant to radiotherapy in the treatmentof patients with GBM. In this regard, a phase II trial isongoing to test GDC-0449 (Vismodegib), an oral small-molecule SMO inhibitor, in patient with recurrent GBM(clinicaltrials.gov; NCT00980343) GDC-0449. This drugwas recently approved by US Food and Drug Adminis-tration for advanced and metastatic basal cell carcinoma(47). In addition Rudin and colleagues have reported thatGDC-0449 induced a rapid but transient regression ofthe tumor in a patient with metastatic medulloblastoma(48). Likewise, our strategy which permits an efficienttargeting of tumoral cells in a necrotic environmentcould be of interest for delivering such type of molecule.

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

Authors' ContributionsConception and design: A. Balbous, B. Renoux, U. Cortes, S. Papot,L. Karayan-TaponDevelopment of methodology: A. Balbous, U. Cortes, K. Guilloteau,T. Legigan, P. Rivet, O. Boissonnade, R.J. Bensadoun, L. Karayan-Tapon

Acquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A. Balbous, U. Cortes, S. Milin, S. Martin,C. Tripiana, M. WagerAnalysis and interpretation of data (e.g., statistical analysis, bio-statistics, computational analysis): A. Balbous, U. Cortes, S. Milin,L. Karayan-TaponWriting, review, and/or revision of themanuscript:A.Balbous,U.Cortes,P. Rivet, O. Boissonnade, S. Martin, M. Wager, R.J. Bensadoun, S. Papot,L. Karayan-TaponAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): A. Balbous, L. Karayan-TaponStudy supervision: R.J. Bensadoun, L. Karayan-Tapon

AcknowledgmentsThe authors thank Adriana Delwail for flow cytometry assistance,

Jeffrey Arsham, an American medical translator, for having reread andreviewed the original English-language text, all the members of theGroupePoitevinde laRecherche contre leCancer (GPRC), andparticularlyProf. J.-M. Muller for fruitful exchange.

Grant SupportThis work was supported by Ligue contre le Cancer de la Vienne et des

Deux-S�evres, R�egion Poitou-Charentes, Canc�eropole Grand Ouest(R�eseau Gliomes), and the "Sport et Collection" and "Rotary Club deCivray" foundations.

The costs of publication of this article were defrayed in part by the pay-ment of page charges. This article must therefore be hereby marked adver-tisement in accordancewith 18U.S.C. Section 1734 solely to indicate this fact.

Received December 5, 2013; revisedMay 19, 2014; accepted June 2, 2014;published OnlineFirst July 22, 2014.

References1. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn

MJB, et al. Radiotherapy plus concomitant and adjuvant temozolo-mide for glioblastoma. N Engl J Med 2005;352:987–96.

2. Galli R, Binda E, Orfanelli U, Cipelletti B, Gritti A, De Vitis S, et al.Isolation and characterization of tumorigenic, stem-like neural pre-cursors from human glioblastoma. Cancer Res 2004;64:7011–21.

3. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, et al.Identification of human brain tumour initiating cells. Nature 2004;432:396–401.

4. Chen R, Nishimura MC, Bumbaca SM, Kharbanda S, Forrest WF,Kasman IM, et al. A hierarchy of self-renewing tumor-initiating celltypes in glioblastoma. Cancer Cell 2010;17:362–75.

5. Beier D, Hau P, Proescholdt M, Lohmeier A, Wischhusen J, Oefner PJ,et al. CD133(þ) and CD133(�) glioblastoma-derived cancer stem cellsshowdifferential growth characteristics andmolecular profiles.CancerRes 2007;67:4010–5.

6. Son MJ, Woolard K, Nam D-H, Lee J, Fine HA. SSEA-1 is an enrich-ment marker for tumor-initiating cells in human glioblastoma. CellStem Cell 2009;4:440–52.

7. Liu G, Yuan X, Zeng Z, Tunici P, Ng H, Abdulkadir IR, et al. Analysis ofgene expressionandchemoresistanceofCD133þcancer stemcells inglioblastoma. Mol Cancer 2006;5:67.

8. KangJ,ShakyaA, TantinD.Stemcells, stress,metabolismandcancer:a drama in two Octs. Trends Biochem Sci 2009;34:491–9.

9. BaoS,WuQ,McLendonRE,HaoY,ShiQ,HjelmelandAB, et al. Gliomastem cells promote radioresistance by preferential activation of theDNA damage response. Nature 2006;444:756–60.

10. Dahmane N, Lee J, Robins P, Heller P, Ruiz i Altaba A. Activation of thetranscription factor Gli1 and the Sonic hedgehog signalling pathway inskin tumours. Nature 1997;389:876–81.

11. Watkins DN, Berman DM, Burkholder SG, Wang B, Beachy PA, BaylinSB. Hedgehog signalling within airway epithelial progenitors and insmall-cell lung cancer. Nature 2003;422:313–7.

12. Sanchez P, Hern�andez AM, Stecca B, Kahler AJ, DeGueme AM,Barrett A, et al. Inhibition of prostate cancer proliferation by interfer-ence with SONIC HEDGEHOG-GLI1 signaling. Proc Natl Acad SciU S A 2004;101:12561–6.

13. BermanDM,KarhadkarSS,MaitraA,MontesDeOcaR,GerstenblithMR,Briggs K, et al. Widespread requirement for Hedgehog ligand stimulationin growth of digestive tract tumours. Nature 2003;425:846–51.

14. Kubo M, Nakamura M, Tasaki A, Yamanaka N, Nakashima H, NomuraM, et al. Hedgehog signaling pathway is a new therapeutic target forpatients with breast cancer. Cancer Res 2004;64:6071–4.

15. Goodrich LV, Milenkovi�c L, Higgins KM, Scott MP. Altered neural cellfates and medulloblastoma in mouse patched mutants. Science1997;277:1109–13.

16. Sandberg CJ, Altschuler G, Jeong J, Strømme KK, Stangeland B,Murrell W, et al. Comparison of glioma stem cells to neural stem cellsfrom the adult human brain identifies dysregulated Wnt- signaling anda fingerprint associated with clinical outcome. Exp Cell Res2013;319:2230–43.

17. Dahmane N, S�anchez P, Gitton Y, Palma V, Sun T, Beyna M, et al. TheSonic Hedgehog-Gli pathway regulates dorsal brain growth andtumorigenesis. Dev Camb Engl 2001;128:5201–12.

18. Bar EE, Chaudhry A, Lin A, Fan X, Schreck K, Matsui W, et al.Cyclopamine-mediated hedgehog pathway inhibition depletesstem-like cancer cells in glioblastoma. Stem Cells 2007;25:2524–33.

19. Clement V, Sanchez P, de Tribolet N, Radovanovic I, Ruiz i Altaba A.HEDGEHOG-GLI1 signaling regulates human glioma growth, can-cer stem cell self-renewal and tumorigenicity. Curr Biol CB 2007;17:165–72.

20. Cooper MK, Porter JA, Young KE, Beachy PA. Teratogen-mediatedinhibition of target tissue response to Shh signaling. Science1998;280:1603–7.

21. Chen JK, Taipale J, Cooper MK, Beachy PA. Inhibition of Hedgehogsignaling by direct binding of cyclopamine to Smoothened. GenesDev2002;16:2743–8.

22. Chen JK, Taipale J, Young KE, Maiti T, Beachy PA. Small moleculemodulation of Smoothened activity. Proc Natl Acad Sci U S A2002;99:14071–6.

23. Kolterud A, Toftgard R. Strategies for Hedgehog inhibition and itspotential role in cancer treatment. Drug Discov Today Ther Strateg2007;229–35.

Balbous et al.





24. Hamon F, Renoux B, Chad�eneau C, Muller J-M, Papot S. Study of acyclopamine glucuronide prodrug for the selective chemotherapy ofglioblastoma. Eur J Med Chem 2010;45:1678–82.

25. RenouxB, Legigan T, Bensalma S, Chad�eneauC,Muller J-M, Papot S.A new cyclopamine glucuronide prodrug with improved kinetics ofdrug release. Org Biomol Chem 2011;9:8459–64.

26. Tranoy-Opalinski I, Legigan T, Barat R, Clarhaut J, ThomasM, RenouxB, et al. b-Glucuronidase-responsive prodrugs for selective cancerchemotherapy: an update. Eur J Med Chem 2014;74:302–13.

27. Nygren C, von Holst H, Ma�nsson JE, Fredman P. Increased activity of

lysosomal glycohydrolases in glioma tissue and surrounding areasfrom human brain. Acta Neurochir 1997;139:146–50.

28. Bosslet K, Straub R, Blumrich M, Czech J, Gerken M, Sperker B, et al.Elucidation of themechanism enabling tumor selective prodrugmono-therapy. Cancer Res 1998;58:1195–201.

29. Juan T-Y, Roffler SR, Hou H-S, Huang S-M, Chen K-C, Leu Y-L,et al. Antiangiogenesis targeting tumor microenvironment syner-gizes glucuronide prodrug antitumor activity. Clin Cancer Res 2009;15:4600–11.

30. Wolburg H, Noell S, Fallier-Becker P, Mack AF, Wolburg-Buchholz K.The disturbed blood-brain barrier in human glioblastoma.Mol AspectsMed 2012;33:579–89.

31. WeissN,Miller F, CazaubonS,CouraudP-O. Theblood-brain barrier inbrain homeostasis and neurological diseases. Biochim Biophys Acta2009;1788:842–57.

32. Wolburg H, Wolburg-Buchholz K, Kraus J, Rascher-Eggstein G, Lieb-ner S, Hamm S, et al. Localization of claudin-3 in tight junctions of theblood-brain barrier is selectively lost during experimental autoimmuneencephalomyelitis and human glioblastoma multiforme. Acta Neuro-pathol 2003;105:586–92.

33. Liu H-L, Hua M-Y, Chen P-Y, Chu P-C, Pan C-H, Yang H-W, et al.Blood-brain barrier disruption with focused ultrasound enhancesdelivery of chemotherapeutic drugs for glioblastoma treatment. Radi-ology 2010;255:415–25.

34. Villalva C,Martin-Lanner�ee S, Cortes U, Dkhissi F,WagerM, LeCorf A,et al. STAT3 is essential for the maintenance of neurosphere-initiatingtumor cells in patients with glioblastomas: a potential for targetedtherapy? Int J Cancer 2011;128:826–38.

35. Villalva C, Cortes U, Wager M, Tourani J-M, Rivet P, Marquant C, et al.O6-Methylguanine-Methyltransferase (MGMT) promoter methylationstatus in glioma stem-like cells is correlated to temozolomide sensi-tivity under differentiation-promoting conditions. Int J Mol Sci 2012;13:6983–94.

36. Balbous A, Cortes U, Guilloteau K, Villalva C, Flamant S, Gaillard A,et al. A mesenchymal glioma stem cell profile is related to clinicaloutcome. Oncogenesis 2014;3:e91.

37. VandenBrinkGR,HardwickJC, TytgatGN,BrinkMA, TenKate FJ, VanDeventer SJ, et al. Sonic hedgehog regulates gastric gland morpho-genesis in man and mouse. Gastroenterology 2001;121:317–28.

38. Van den Brink GR, Bleuming SA, Hardwick JCH, Schepman BL,Offerhaus GJ, Keller JJ, et al. Indian Hedgehog is an antagonist ofWnt signaling in colonic epithelial cell differentiation. Nat Genet2004;36:277–82.

39. TakebeN,Harris PJ,WarrenRQ, IvySP. Targeting cancer stemcells byinhibiting Wnt, Notch, and Hedgehog pathways. Nat Rev Clin Oncol2011;8:97–106.

40. Berman DM, Karhadkar SS, Hallahan AR, Pritchard JI, Eberhart CG,Watkins DN, et al. Medulloblastoma growth inhibition by hedgehogpathway blockade. Science 2002;297:1559–61.

41. Sanchez P, Ruiz i Altaba A. In vivo inhibition of endogenous braintumors through systemic interference of Hedgehog signaling in mice.Mech Dev 2005;122:223–30.

42. Feldmann G, Habbe N, Dhara S, Bisht S, Alvarez H, Fendrich V, et al.Hedgehog inhibition prolongs survival in a genetically engineeredmouse model of pancreatic cancer. Gut 2008;57:1420–30.

43. Lipinski RJ, Hutson PR, Hannam PW, Nydza RJ, Washington IM,Moore RW, et al. Dose- and route-dependent teratogenicity, toxicity,and pharmacokinetic profiles of the hedgehog signaling antagonistcyclopamine in the mouse. Toxicol Sci Off J Soc Toxicol 2008;104:189–97.

44. Ruiz i Altaba A, Mas C, Stecca B. The Gli code: an information nexusregulating cell fate, stemness and cancer. Trends Cell Biol 2007;17:438–47.

45. Beier D, R€ohrl S, Pillai DR, Schwarz S, Kunz-Schughart LA, Leukel P,et al. Temozolomide preferentially depletes cancer stem cells inglioblastoma. Cancer Res 2008;68:5706–15.

46. Legigan T, Clarhaut J, Renoux B, Tranoy-Opalinski I, Monvoisin A,Berjeaud J-M, et al. Synthesis and antitumor efficacy of a b-glucu-ronidase-responsive albumin-binding prodrug of doxorubicin. J MedChem 2012;55:4516–20.

47. Sekulic A, MigdenMR, Oro AE, Dirix L, Lewis KD, Hainsworth JD, et al.Efficacy and safety of vismodegib in advancedbasal-cell carcinoma.NEngl J Med 2012;366:2171–9.

48. Rudin CM, Hann CL, Laterra J, Yauch RL, Callahan CA, Fu L, et al.Treatment ofmedulloblastomawith hedgehog pathway inhibitor GDC-0449. N Engl J Med 2009;361:1173–8.






2014;13:2159-2169. Published OnlineFirst July 22, 2014.Mol Cancer Ther Anaïs Balbous, Brigitte Renoux, Ulrich Cortes, et al. Stem-like Cancer Cell Inhibition in GlioblastomaSelective Release of a Cyclopamine Glucuronide Prodrug toward

Updated version

10.1158/1535-7163.MCT-13-1038doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2014/07/23/1535-7163.MCT-13-1038.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/13/9/2159.full#ref-list-1

This article cites 47 articles, 13 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mct.aacrjournals.org/content/13/9/2159To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-13-1038

http://mct.aacrjournals.org/content/suppl/2014/07/23/1535-7163.MCT-13-1038.DC1

http://mct.aacrjournals.org/content/13/9/2159.full#ref-list-1

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

mailto:[email protected]

http://mct.aacrjournals.org/content/13/9/2159


selective release of a cyclopamine glucuronide prodrug toward stem … · cyclopamine has inhibited...

Documents