PTEN Loss Mitigates the Response of …...Medulloblastoma is the most common malignant pediat-ric...

Therapeutics, Targets, and Chemical Biology

PTEN Loss Mitigates the Response of Medulloblastoma toHedgehog Pathway Inhibition

Ciara Metcalfe, Bruno Alicke, Ailey Crow, Marlea Lamoureux, Gerrit J.P. Dijkgraaf, Franklin Peale,Stephen E. Gould, and Frederic J. de Sauvage

AbstractMedulloblastoma is a cancer of the cerebellum, for which there is currently no approved targeted therapy.

Recent transcriptomics approaches have demonstrated that medulloblastoma is composed of molecularlydistinct subgroups, one of which is characterized by activation of the Hedgehog pathway, which in mousemodels is sufficient to drive medulloblastoma development. There is thus considerable interest in targetingthe Hedgehog pathway for therapeutic benefit in medulloblastoma, particularly given the recent approval ofthe Hedgehog pathway inhibitor vismodegib for metastatic and locally advanced basal cell carcinoma. Likeother molecularly targeted therapies, however, there have been reports of acquired resistance to vismodegib,driven by secondary Hedgehog pathway mutations and potentially by activation of the phosphatidylinositol3-kinase (PI3K) pathway. Given that acquired resistance to vismodegib may occur as a result of inappropriatePI3K pathway activation, we asked if loss of the PI3K pathway regulator, phosphatase and tensin homologue(Pten), which has been reported to occur in patients within the Hedgehog subgroup, would constitute amechanism of innate resistance to vismodegib in Hedgehog-driven medulloblastoma. We find that Hedgehogpathway inhibition successfully restrains growth of Pten-deficient medulloblastoma in this mouse model, butdoes not drive tumor regression, as it does in Pten-wild-type medulloblastoma. Combined inhibition of theHedgehog and PI3K pathways may lead to superior antitumor activity in PTEN-deficient medulloblastoma inthe clinic. Cancer Res; 73(23); 7034–42. �2013 AACR.

IntroductionMedulloblastoma is the most common malignant pediat-

ric brain tumor, and is the leading cause of cancer-relatedmortality in childhood. Current available therapy consists ofsurgery, radiation therapy of the brain and spinal cord,followed by chemotherapy. Unfortunately, this aggressivetherapeutic strategy is associated with serious adverseeffects, including postoperative mutism, neurocognitive def-icits, endocrinopathies, and increased susceptibility to sec-ondary malignancies (1).

Recent transcriptional profiling has demonstrated thatmedulloblastoma is composed of 4 molecularly distinct sub-groups: theWnt subgroup, Sonic Hedgehog subgroup, Group 3,andGroup 4 (2). The identification of these subgroups providesamuch needed opportunity for the use of molecularly targetedtherapeutics in medulloblastoma, which have the potential toboth increase efficacy and reduce side effects when comparedwith the current standard of care. The Hedgehog subgroup,

which makes up approximately 25% of medulloblastoma cases(3), may be particularly amenable to a targeted therapeuticapproach, given the recent U.S. Food and Drug Administrationapproval of the Hedgehog pathway inhibitor, vismodegib(previously known as GDC-0449), in metastatic and locallyadvanced basal cell carcinoma (BCC).

Hedgehog signaling is critical for embryonic developmentand also plays a limited role in adult homeostasis. In the off-state, the transmembrane receptor Patched (PTCH) sup-presses the activity of the GPCR-like molecule Smoothened(SMO). The Hedgehog ligands (Sonic, Indian, or Desert Hedge-hog) activate the pathway by binding to PTCH, which releasesits inhibition over SMO. SMO in turn signals to theGLI family oftranscription factors and thus induces a change in the tran-scriptional profile of the cell (4). Inappropriate activation of thepathway, most often caused by inactivating mutations inPTCH, or activating mutations in SMO, is associated withcancer, predominantly BCC and medulloblastoma (5–8).

Vismodegib binds to, and inhibits SMO, blockingdownstream Hedgehog pathway activation

In early-stage clinical studies, a patient with medulloblas-toma from the Hedgehog subgroup harboring widespreadmetastatic disease, exhibited rapid and dramatic tumor regres-sion when treated with vismodegib (9). Unfortunately, therobust initial response was followed by relapse, reminiscentof the acquired resistance seen in the context of other targetedtherapies, such as erlotinib (10). Sequence analysis of a tumor

Authors' Affiliation: Genentech Inc., South San Francisco, California

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

Corresponding Author: Frederic J. de Sauvage, Genentech Inc., 1 DNAWay, South San Francisco 94080, CA. Phone: 650-225-5841; Fax: 650-225-6497. E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-1222

�2013 American Association for Cancer Research.

CancerResearch

Cancer Res; 73(23) December 1, 20137034

on August 7, 2020. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 23, 2013; DOI: 10.1158/0008-5472.CAN-13-1222

http://cancerres.aacrjournals.org/

biopsy obtained after progression in this patient indentified anamino acid substitution at a conserved residue of SMO. TheD473H substitution disrupts the ability of vismodegib to bindSMO, but does not alter the ability of SMO to activate theHedgehog pathway (11). This observation triggered a series ofinvestigations into potential alternative mechanisms of resis-tance against vismodegib and sonidegib, another Hedgehogpathway inhibitor currently under clinical development (12). Avariety of resistance mechanisms were identified through theuse of mouse models, including additional mutations in Smo,as well as downstream Hedgehog pathway alterations such asGli2 amplification (11, 13, 14). An additional, and unexpectedpotential mechanism of acquired resistance to sonidegib wasthe upregulation of PI3K signaling (13).The PI3K pathway is a well-established oncogenic pathway

that regulates cell growth, proliferation, and survival by relay-ing growth factor signaling (15). Aberrant activation of thePI3K pathway occurs in multiple tumor types and is frequentlydriven by disruptions in the negative pathway regulator PTEN,but can also be driven by gain-of-function mutations inPIK3CA. A recent genomic analysis ofmedulloblastoma tumorsrevealed that of 13 Hedgehog subgroup patients profiled, 2 hadloss-of-function mutations in PTEN, and another patienthad an activating mutation in PIK3CA (16). Of 66 patientsprofiled from the other subgroups, none had loss of PTEN, and2 had mutations in PIK3CA. Another sequencing study simi-larly found a number of PTEN mutations in medulloblastomatumors, one of which co-occurred with a homozygous PTCHmutation (17). In addition to genomic alterations as a mech-anism to hyperactivate the PI3K pathway, epigenetic inacti-vation of PTEN has been reported to occur at a high frequencyin medulloblastoma samples (18). Given that inappropriateactivation of PI3K signaling may represent a potential mech-anism of acquired resistance to Hedgehog pathway inhibitors,we wanted to determine if preexisting PTENmutations, whichoccur in Hedgehog subgroup patients, would alter the initialresponse to vismodegib in Hedgehog-drivenmedulloblastoma.

Materials and MethodsMouse modelsMath1CreER and Ptenloxp mouse strains were both obtained

from The Jackson Laboratory (stock numbers 007684 and004597, respectively), and the Ptchloxp strain was a kind giftfrom R. Toftgard and S. Teglund (Karolinska Institutet, Stock-holm, Sweden; ref. 19). All mice were housed and maintainedaccording to the animal-use guidelines of Genentech, con-forming to California State legal and ethical practices. Cre-dependent recombination was induced by oral gavage oftamoxifen (single dose of 4 mg in sunflower seed oil) topregnant dams at E14.5.

Tissue analysisTumors were harvested frommoribund animals and subject

either to overnight fixation in 4% paraformaldehyde beforeembedding in paraffin, or to lysis in CelLytic MT Lysis reagent(Sigma). Immunohistochemistry (IHC), immunofluorescence,and Western blotting were carried out according to standardprocedures. Antibodies used were as follows: mouse anti-NeuN

(1:500 for immunofluorescence and 1:1000 forWesternblotting;Millipore, #MAB377), rabbit anti-Ki67 (1:300; ThermoScientific,#RM9106), rabbit anti-pAKT S473 (1:500 for IHC and 1:1000 forWestern blotting; Cell Signaling Technology, #4060), rabbitanti-pS6 (1:1000; Cell Signaling Technology, #4858), rabbitanti-PTEN (1:100 for IHC and 1:1000 for Western blotting; CellSignalingTechnology, #9188), and rabbit anti-cleaved caspase-3(1:100 for IHC; Cell Signaling Technology, #9661). The DAKOEnvisionþ horseradish peroxidase (DAB) system was used todetect primary antibodies for IHC. Images were acquired usinga Zeiss Axioskop2 plusmicroscopefittedwith anAxioCamHRcor a Leica DMI4000B fluorescence microscope. Adobe Photo-shop CS3 was used to process images, and Fiji was used toquantify cleaved caspase-3 staining.

Generation of medulloblastoma allografts and drugtreatments

To enable an assessment of the sensitivity of these medul-loblastoma models to inhibition of the Hedgehog and PI3Kpathways, tumor growth was relocated subcutaneously viagrafting. The cerebella of mice displaying symptoms associ-ated with medulloblastoma were harvested and prepared forsubcutaneous inoculation into the right lateral thorax offemale CD1-nude (CRL) mice aged at least 6 weeks, toestablish passage 1. Allografts were generated via inoculationof a single cell suspension in 100 mL Neuralbasal Medium(Invitrogen; PM model, derived from Ptchloxp/loxp; Math1CreER

mice) or small tumor fragments (PPM model, derived fromPtchloxp/loxp; Ptenloxp/loxp; Math1CreER mice). Serial in vivo prop-agation of the PPM model was repeated to generate sufficienttumor-bearing animals for drug treatment. Tumors weremeasured with calipers, and tumor volumes calculated usingthe formula v ¼ 0.5 � length � width2, where length andwidth represent perpendicular tumor diameters. As tumorsreached 100 mm3, tumor-bearing animals were separated intogroups of similarly sized tumors and drug administration wasinitiated.

Compounds were formulated in 0.5% methylcellulose/0.2%Tween 80 (MCT) and mice were administered either vehicle ordrug once daily (qd) by oral gavage; vismodegib and GNE-317treatments were separated by 4 hours. To generate samples forpharmacodynamic analysis (quantitative PCR and Westernblotting), tumor-bearing mice were treated for 3 consecutivedays, and tumors were harvested 4 hours after the final dose.

Statistical methodsAmixedmodeling approach was used to analyze the repeat-

edmeasurement of tumor volumes from the same animals overtime (20). This approach addresses both repeated measure-ments andmodest dropouts before study end. Restricted cubicsplines were used to fit a nonlinear profile to the time coursesof log2(tumor volume) in each group. Fitting was done via alinear mixed effects model, using the R package nlme, version3.1 97 in R version 2.12.0 (R Development Core Team 2008; RFoundation for Statistical Computing).

Tumor growth inhibition (TGI) as a percentage of vehiclecontrol (%TGI) was calculated as the percentage of the areaunder the fitted tumor volume–time curve (AUC) on the linear

Hedgehog Pathway Inhibition in PTEN-Deficient Medulloblastoma

www.aacrjournals.org Cancer Res; 73(23) December 1, 2013 7035




scale for the respective treatment group per day in relation tothe starting volume on Day 0 (baseline) and in relation to thevehicle. As such, a TGI value of 100% indicates tumor stasis; aTGI value of >1% but <100% indicates tumor growth delay; anda TGI value of >100% indicates tumor regression and would beassociated with a negative AUC.

To determine uncertainty intervals (UI) for %TGI, thefitted curve and the fitted covariance matrix were used togenerate a random sample as an approximation to thedistribution of %TGI. The random sample is composed of1,000 simulated realizations of the fitted mixed model, inwhich the %TGI is recalculated for each realization. Ourreported UI is the value for which 95% of the time therecalculated values of %TGI will fall in this region given thefitted model. The 2.5 and 97.5 percentiles of the simulateddistribution were used as the upper and lower UIs, respec-tively. All TGI values were calculated at day 13 of theexperiment, to allow for direct comparison across the twomodels; TGI cannot be calculated beyond day 13 in the PPMmodel as control/vehicle-treated tumor-bearing animalswere euthanized before the end of the study, because tumorvolumes exceeded 2,000 mm3.

ResultsLoss of PTEN accelerates the onset of morbidity in amouse model of Hedgehog-driven medulloblastoma

An earlier study showed that deletion of Ptch in cerebellargranule neural precursor cells (CGNP), which was achieved bycrossing Math1CreER mice to Ptchloxp/loxp mice, resulted in thedevelopment of medulloblastoma, with 100% penetrance (21).For our studies, we combined this model with the previouslydescribed Ptenloxp/loxp mouse allele to generate Ptchloxp/loxp;Ptenloxp/loxp; Math1CreER compound mutants (22). Similar tothe previous report, Ptchloxp/loxp;Math1CreER (hereafter referredto as PM) mice (induced with tamoxifen at E14.5) display signsof illness during early adulthood (6–10 weeks), which include adomed head and abnormal gait. Strikingly, loss of Pten inPtchloxp/loxp; Ptenloxp/loxp; Math1CreER (referred to as PPM) micedramatically accelerates disease, with animals as young as 3weeks of age presenting with ataxia; the median survival is 9.3weeks in PM versus 4.0 weeks in PPM mice. Ptchloxp/loxp;Ptenwt/loxp; Math1CreER (PPwt/loxpM) littermates have an inter-mediate survival, with a median of 7.4 weeks (Fig. 1A). Impor-tantly, loss of Pten alone in this model is not sufficient toinitiate observable disease, as Ptchwt/wt; Ptenloxp/loxp; Math1CreER

mice survive long term with no symptoms.Analysis of brain samples from PM and PPM mice

revealed very large tumors within the cerebellum, whichdisrupt the normal cerebellar architecture (Fig. 1B–D). His-tologically, tumors from PM mice resemble classic medul-loblastoma, which is the histological subtype that predomi-nates among human patients from the Hedgehog subgroup;these are frequently described as "small round blue-cell"tumors (Fig. 1C). In contrast, tumors from PPM mice have amarkedly different appearance, with a histological pheno-type analogous to medulloblastoma with extensive nodular-ity (MBEN), which in humans, is a histological subtypethought to be restricted to Hedgehog subgroup patients

(3). Tumors from PPwt/loxpM mice are histologically indis-tinguishable from PPM tumors, with an MBEN appearance(Supplementary Fig. S1). The cerebella of Ptchwt/wt;Ptenloxp/loxp; Math1CreER mice lack tumors, however, we diddetect large ectopic cells both external to, and within themolecular layer (Supplementary Fig. S2).

Figure 1. Loss of Pten in Hedgehog-driven medulloblastomaaccelerates tumor onset and alters tumor histology. A, animals displayinghydrocephalus with clear ataxia were humanely euthanized, and age atsacrifice was used to generate a Kaplan–Meier survival curve. Ptchwt/wt;Ptenloxp/loxp; Math1CreERmice do not succumb to disease and are still alivea year after birth; those with loss of Ptch, including Ptchloxp/loxp; Ptenwt/wt;Math1CreER (PM), Ptchloxp/loxp; Ptenwt/loxp; Math1CreER and Ptchloxp/loxp;Ptenloxp/loxp; Math1CreER (PPM), all show severe clinical symptoms before20 weeks of age. B–D, hematoxylin and eosin staining of a normalcerebellum (B) and cerebellar tumors from PM (C) and PPM (D) mice. Notethe distinct appearance of the PM tumor bulk, with very tightly packednuclei and almost no eosinophilic material versus the PPM tumor with lessdensely packednuclei separatedbyeosinophilic tissue.Granule cells in thepresumptive IGL are outlined. Scale bars represent 1 mm in the low-magnification images and 100 mm in the high magnification images.

Metcalfe et al.

Cancer Res; 73(23) December 1, 2013 Cancer Research7036




Activation of PI3K signaling drives medulloblastomacells toward a more differentiated phenotypeImmunohistochemical analysis revealed additional differ-

ences between PM and PPM tumors. PM tumors are largelymade up of undifferentiated cells, positive for Ki67 and neg-ative for the neuronal marker NeuN (Fig. 2A, see also Supple-mentary Fig. S3 for phospho-histone H3 IHC). Somewhatcounterintuitively, the more aggressive PPM tumors have ahigh content of differentiated cells, positive for NeuN andnegative for Ki67, with Ki67þ cells occurring sporadicallywithin the tumor mass. Strikingly, in medulloblastoma tumorsthat are eitherPtenloxp/loxp orPtenloxp/wt, in addition to scatteredKi67 positive cells, we also find regions of Ki67þ/NeuN� cellsthat occur as "stripes" (Fig. 2B, and Supplementary Fig. S3 for

pH3 IHC). These proliferative stripes seem to localize toperivascular regions (see Fig. 2C, inset), an observation thatwas confirmed by EdU/CD31 double staining, in tumors har-vested 2 hours after administration of the thymine analog (Fig.2D). PM tumors have a relatively high level of baseline apo-ptosis, as revealed by the presence of cleaved caspase-3(CC3), which seemed to be markedly reduced in PPM tumors(Fig. 2E and F). We used image analysis to quantify thisapparent difference, taking into account the observation thatin any field of view there seem to bemore cells in the PMmodelthan in the PPM model, and found that indeed, apoptosis issignificantly suppressed in the PPM model (Fig. 2G, see alsoSupplementary Fig. S4 for an illustration how the imageanalysis was conducted).

Figure 2. Activation of the PI3Kpathway promotes differentiationin the context of Hedgehog-drivenmedulloblastoma. A, PM tumorshave a consistent pattern of Ki67positivity and lack of differentiationthroughout the tumor bulk. NeuN iscaptured in the green channel andKi67 in red. Differentiated granulecells of the IGL are outlined, and anasterisk indicates the tumor bulk.B, PPM tumors, as well as thosefrom Ptchloxp/loxp; Ptenwt/loxp;Math1CreER mice (C), stainpositive for NeuN throughout thetumor bulk, with Ki67 positivityoccurring sporadically within thetumor bulk and along putativeblood vessels. Note the presenceof green, auto-fluorescent redblood cells within the proliferativestripe, highlighted in the inset bya white arrow. D, EdU, an analogof BrdU, was administered toPtchloxp/loxp; Ptenwt/loxp;Math1CreER animals 2 hours beforeharvest to enable the identificationof actively cycling cells. EdUpositive cells (green channel) seemto be preferentially clustered nearvessels, which stain positive forCD31 (red channel). E and F, IHCfor cleaved caspase-3 (CC3) in atumor section from a PMmouse (E)and from a PPM mouse (F). G,quantification of CC3 staining inPM and PPM tumor sections;individual dots represent % CC3positive area/total nuclear areawithin a single image; 7 to 8 imageswere analyzed per animal (n ¼ 3,P < 0.0001). Error bars, SEM. Scalebars, 100 mm.






Loss of PTEN in Hedgehog-driven medulloblastomaalters the response of tumors to vismodegib

The early onset of tumors in the PM and PPM models, thevery rapid onset of morbidity of the PPM model in particular,and the lack of normal cerebellar tissue in this model, whichwould be required to maintain cerebellar functionality aftertargeting the tumor, prevented us from performing a vismo-degib intervention study in the autochthonous setting. Fortu-nately however, the value and relevance of allograft studies inthe context of medulloblastoma have previously been wellestablished (11, 13, 14). Therefore, to evaluate the response ofPM and PPM tumors to vismodegib, we generated allograftmodels; tumors harvested from PM and PPM animals werepassaged subcutaneously in nude mice. This allowed trackingof tumor volumes over time, in response to various doses ofdrug (see also Materials and Methods). Although the histolog-ical characteristics distinguishing primary PM and PPMtumors do not seem to be strictly maintained in the allograftsetting, key features, such as PTEN status and pAKT levels (Fig.3A and B), aremaintained. Note the loss of PTEN in PPM tumorcells, but not in stromal cells, consistent with the restriction ofcre-recombinase activity to Math1-expressing neural progeni-tors. As expected, loss of PTEN is associated with robustupregulation of pAKT. In addition, PPM allografts seem toretain a somewhat more differentiated phenotype when com-pared with PM allografts, with apparent increases in NeuNlevels (Fig. 3C), although the differentiation phenotype is lessdramatic in the allografts when compared with the primarytumor.

Oral gavage of mice with 3 mg/kg vismodegib qd has littleeffect on allograft growth, in either model, despite having amodest effect onHedgehog pathway activity (Fig. 4A, B, andD).This finding, that �50% Hedgehog pathway inhibition is notsufficient to drive a tumor response, is consistent with therecent demonstration that sustained and robust inhibition ofthe Hedgehog pathway (>80%) is required for a meaningful

antitumor effect (23). In contrast, 30 mg/kg vismodegib qd issufficient to cause a meaningful growth delay in both models,with a mean TGI of 90% in the PMmodel, and 70% in the PPMmodel (Fig. 4A–C).

Interestingly, the response to vismodegib seems to divergebetween the models at higher doses of vismodegib. In the PMmodel, vismodegib dosed at 60 and 90mg/kg qd results in rapidand robust tumor regression, with 149% and 159% mean TGI,respectively; tumors continued to regress beyond day 13, andwere barely detectable at day 21, when the study ended (Fig. 4Aand C). In contrast, PPM tumors had not regressed by day 13,on either 60 or 90 mg/kg vismodegib, with a mean TGI of lessthan 100% in both cases (87% and 90%); rather, these dosesachieved tumor stasis by the end of the study, despite acomparable level of Hedgehog pathway inhibition (Fig. 4D).

CombinedPI3KandHedgehogpathway inhibition drivestumor regression in the PPM model

GNE-317 is a PI3K/mTOR pathway inhibitor that wasspecifically designed to cross the blood–brain barrier, withthe treatment of PI3K pathway-driven glioblastoma as theprimary objective (24). We made use of this previouslydescribed compound to ask if inhibition of the Hedgehogpathway in combination with inhibition of PI3K signaling,would improve on vismodegib as a single agent. GNE-317when dosed at 30 mg/kg qd successfully attenuates PI3Ksignaling in the allograft setting, as measured by pS6 levels(Fig. 5A), and by pAKT levels (Supplementary Fig. S5), butdoes not alter levels of Hedgehog pathway activity (Fig. 4D).Similarly, 90 mg/kg vismodegib does not influence activationof the PI3K pathway in this setting (Supplementary Fig. 5).GNE-317, as a single agent, achieved modest inhibition oftumor growth in the PM model, and a more meaningfulinhibition of growth in the PPM model (mean TGI 37% and66%, respectively; Fig. 5B–D). In the PPM model, althoughvismodegib treatment alone did not drive regression at any

Figure 3. PM and PPM allograftsreflect key features of the primarytumors. A, hematoxylin and eosinstaining of PM and PPM allograftmodels. B, IHC for PTEN and pAKTin primary and allograft models. C,Western blot analysis of lysatesgenerated from three vehicle-treated PM and PPM allografts.Note that NeuN is known to occuras multiple isoforms of distinctmolecular weights (34). Scale bars,100 mm.

Metcalfe et al.





concentration tested, GNE-317 treatment, when combinedwith either 60 or 90 mg/kg vismodegib, resulted in tumorregression (Fig. 5B and D). At 60 mg/kg vismodegib plusGNE-317, mean tumor volume at the study start was 232mm3, which regressed to 152 mm3 on day 13 and 100 mm3 bythe study endpoint. At 90 mg/kg vismodegib plus GNE-317,mean tumor volume at the study start was 244 mm3, whichregressed to 102 mm3 on day 13 and 80 mm3 by the studyend. It is also worth noting that the regression driven by thecombination of GNE-317 plus 90 mg/kg vismodegib is highlyconsistent across individual tumors/animals (Fig. 5D).

DiscussionInherited mutations in PI3K signaling components are

associated with human syndromes that include neurologicalabnormalities, such as Lhermitte–Duclos disease, indicatingthat this pathway is important in normal neural developmentand/or function (25). Indeed, 2 independent mouse studiesdescribed the use of GfapCre to delete Pten in the brain, andshowed that this led to lethal neurological defects includingseizures and ataxia, which were coupled to an enlargement ofthe cerebellum (25, 26). Histological analysis revealed the

presence of ectopic, enlarged cells at the pial surface andwithin the molecular layer of the cerebellum, which wasattributed to a migration defect in mutant cells. Histologically,this phenomenon is very similar to the phenotype seen inPtchwt/wt; Ptenloxp/loxp; Math1CreER mice, although the micedescribed here do not display any overt clinical symptoms,potentially because of incomplete deletion of PTEN driven bythe Math1CreER (Supplementary Fig. S2). A more recent studydemonstrated that postnatal deletion of PTEN in migratingneuroblasts in the rostral migratory stream resulted in ectopicpositioning and altered morphology of neurons. The authorsargue that the migration defect associated with PTEN loss waslikely secondary to precocious differentiation, rather than adefect in the mechanics of directional migration (27). Thehighly differentiated nature of the tumors in PPM micedescribed here, as well as NeuN expression in ectopic cells inthe cerebella of Ptchwt/wt; Ptenloxp/loxp; Math1CreER mice, sup-ports the notion that loss of PTEN may drive precocious/premature differentiation in neurons, although the presence ofPTEN-deficient neurons within the inner granule layer (IGL)argues against a migration defect of these cells. Perhaps mostrelevant, the RCAS/tv-a system, which allows postnatal genetransfer in a cell-type-specific manner, was used to generate

Figure 4. Hedgehog pathwayinhibition causes robust regressionin the PM allograft model andtumor stasis in the PPM model. A,dose response of vismodegib onPM tumor growth. Doses ofvismodegib are displayed on thetop horizontal gray bar as mg/kg.Gray lines show tumor volume overtime for individual animals (1 tumorper animal). The dashed blue linesshow the tumor growth fit forvehicle-treated tumors, whereasthe black line shows tumor growthfit for tumors at the indicated dosesof vismodegib. Solid red linesindicate tumor volume traces foranimals that were euthanizedbefore study end due to largetumors. B, tumor volumeplots fromPPM animals at various doses ofvismodegib as above. C, TGI (%)driven by indicated doses ofvismodegib at day 13 of theexperiment, compared withvehicle-treated control tumors. D,expression analysis (quantitativePCR) of the Hedgehog target gene,Hedgehog-interacting protein(HIP), in PM and PPM samplestreated as indicated, normalizedagainst expression of thehousekeeping gene Rpl19, withMCT-treated samples being set to100%, n ¼ 3. GNE-317 is a PI3K/mTOR pathway inhibitor. Note thatGNE-317 does not suppressHedgehog pathway activation.Error bars, SEM.






medulloblastomaby targeting SHH tonestin-expressing neuralstem cells, either in the presence or absence of PTEN (28).Intriguingly, PTEN-deficient tumors were described as havingMBEN histology, with high levels of NeuN, very similar to thePPMmodel described here.Moreover, the perivascular niche ofthese tumors was described as being highly proliferative, "infact the only proliferative region in the tumors," again, verysimilar to the PPM model described here. We propose thatthese proliferative cells in the perivascular niche are majorcontributors to tumor growth. A different study, using aconstitutively active mutant of Smo, SmoA1, to induce medul-loblastoma in Ptenþ/� mice similarly demonstrated that acti-vation of PI3K signaling accelerated tumorigenesis while driv-ing a switch from classic to MBEN histology, and concomi-tantly increased neuronal differentiation (29). Notably though,although there is a clear relationship between hyperactivationof the PI3K pathway and MBEN histology that is recapitulatedin a number of different mouse studies, it remains to beestablished if the same relationship exists in human patients.

That activation of PI3K signaling can simultaneously pro-mote differentiation while also accelerating tumorigenesis issomewhat surprising; differentiated tumors are generally asso-ciated with slower growth and a better prognosis. Intriguingly

though, blocking apoptosis in a SmoA1 model of medulloblas-toma, by deleting proapoptotic Bax resulted in a very similarphenotype: a paradoxical increase in differentiation coupled toacceleration of disease (30). Strikingly, SmoA1; Bax�/� tumorsseem indistinguishable from the PPM tumors described here(see Fig. 5, Garcia et al.). The authors propose amodel in whichmedulloblastoma cells face a cell fate choice between apopto-sis and differentiation; those cells that are competent toundergo apoptosis will do so and will thus be lost from thetumor, resulting in tumor turnover. In contrast, cells incapableof launching an apoptotic program will adopt a terminallydifferentiated fate, leading to the accumulation of differenti-ated cells. The differentiated, nonapoptotic, and aggressivephenotype of the PPMmodel described here, together with theestablished role of PI3K signaling in promoting cell survival, isin line with this fate choice model.

Despite the profound changes that occur in PTEN-deficienttumors with respect to PTENwild-type tumors, PPM allograftsare responsive to vismodegib: inhibition of the Hedgehogpathway restrains tumor growth and results in stasis, suggest-ing that even in the context of PTEN loss, tumors remain highlydependent on Hedgehog pathway activity for growth. Trulyresistant tumors, for example allografts that harbor the mouse

Figure 5. A combination approachresults in regression of PPMtumors. A, Western blot analysis ofallograft lysates treated as shown,demonstrating attenuation of PI3Kpathway activity by GNE-317. B,TGI driven by a fixed dose of30 mg/kg GNE-317 plus indicateddoses of vismodegib at day 13of the experiment, compared withvehicle-treated control tumors. C,individual tumor volume plotsfrom PM allografts. Doses ofvismodegib (mg/kg) are shownacross the top gray bar, with a fixeddose of the PI3K pathway inhibitor.D, individual tumor volume tracesfrom PPM allografts as above.

Metcalfe et al.





equivalent of the patient-derived SMO D473H mutationdescribed above, continue to grow robustly in the presenceof 75 mg/kg vismodegib (11). Importantly then, althoughhyperactivation of PI3K signaling prevents regression oftumors in response to Hedgehog pathway inhibition, it doesnot confer bona fide resistance. Nevertheless, the distinctionbetween stasis and regression is of critical importance, andraises a key question: how is regression driven by Hedgehogpathway inhibition in the PMmodel, and how is this altered bythe loss ofPten? Another open question relates to the durabilityof the drug response in these models; specifically how thedurability is influenced by the lack of regression in vismodegib-treated Pten mutant tumors. It is likely that in the absence ofcomplete regression, discontinuation of drug treatment willlead to immediate reinitiation of tumor growth, as occurs inPtchþ/�; p53�/� allograft tumors (11).Our findings have significant implications for the clinical

development of Hedgehog pathway inhibitors in medulloblas-toma, which are currently ongoing (see clinicaltrials.gov, andrefs. 31–33) for descriptions of Hedgehog pathway inhibitorscurrently under clinical development). In the case of BCC, forwhich vismodegib is nowapproved, the vastmajority of tumorsdisplay alterations in the Hedgehog pathway, and there is nocurrent requirement for the stratification of patients. However,the additional genomic complexity seen in patients withmedulloblastoma means that it will be critical to preselectthe �25% of patients who are defined as the Hedgehogsubgroup. Based on the data described here, we now arguethat it will also be important to monitor PI3K pathway statusduring patient selection, particularly in cases where a positiveclinical trial outcome is defined by regression, rather thanstasis. The diagnostic strategy, which will enable the selectionof patients who may respond to Hedgehog pathway inhibitors

as a single agent, versus those who might benefit from acombined regime of a Hedgehog pathway inhibitor plus abrain penetrant PI3K inhibitor, will require careful consider-ation, given that PI3K pathway alterations inmedulloblastomahave been reported not only at the genomic, but also at theepigenetic level (16, 18). It will also be critical to determine ifand how other alterations that have been reported to occur inthe Hedgehog subgroup, such asN-myc amplification, will alterthe response of Hedgehog-driven medulloblastoma to Hedge-hog pathway inhibition.

Disclosure of Potential Conflicts of InterestC. Metcalfe is employed as a postdoctoral fellow in Genentech Inc. A. Crow is

employed as a senior research associate in Genentech Inc. F.V. Peale is employedas a senior pathologist in Genentech, Inc. F.V. Peale also has ownership interest(including patents) in Roche. F.J. de Sauvage has ownership interest (includingpatents) in Roche. No potential conflicts of interest were disclosed by the otherauthors.

Authors' ContributionsConception and design: C. Metcalfe, F.J. de SauvageDevelopment of methodology: B. Alicke, S.E. GouldAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Metcalfe, B. Alicke, M. LamoureuxAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Metcalfe, B. Alicke, A. Crow, F.V. Peale, F.J.de SauvageWriting, review, and/or revision of the manuscript: C. Metcalfe, B. Alicke,G.J.P. Dijkgraaf, S.E. Gould, F.J. de SauvageAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): F.J. de SauvageStudy supervision: S.E. Gould, F.J. de Sauvage

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

Received April 29, 2013; revised July 29, 2013; accepted September 23, 2013;published OnlineFirst October 23, 2013.

References1. Polkinghorn WR, Tarbell NJ. Medulloblastoma: tumorigenesis, cur-

rent clinical paradigm, and efforts to improve risk stratification. NatClin Pract Oncol 2007;4:295–304.

2. TaylorMD,Northcott PA, KorshunovA, RemkeM,ChoYJ, Clifford SC,et al. Molecular subgroups of medulloblastoma: the current consen-sus. Acta Neuropathol 2012;123:465–72.

3. Northcott PA, Korshunov A, Pfister SM, Taylor MD. The clinicalimplications of medulloblastoma subgroups. Nat Rev Neurol 2012;8:340–51.

4. Varjosalo M, Taipale J. Hedgehog: functions and mechanisms. GenesDev 2008;22:2454–72.

5. Epstein EH. Basal cell carcinomas: attack of the hedgehog. Nat RevCancer 2008;8:743–54.

6. Pietsch T, Waha A, Koch A, Kraus J, Albrecht S, Tonn J, et al.Medulloblastomas of the desmoplastic variant carry mutations of thehuman homologue of Drosophila patched. Cancer Res 1997;57:2085–8.

7. Raffel C, Jenkins RB, Frederick L, Hebrink D, Alderete B, Fults DW,et al. Sporadic medulloblastomas contain PTCH mutations. CancerRes 1997;57:842–5.

8. Wolter M, Reifenberger J, Sommer C, Ruzicka T, Reifenberger G.Mutations in the human homologue of theDrosophila segment polaritygene patched (PTCH) in sporadic basal cell carcinomas of the skin andprimitive neuroectodermal tumors of the central nervous system.Cancer Res 1997;57:2581–5.

9. Rudin CM, Hann CL, Laterra J, Yauch RL, Callahan CA, Fu L, et al.Treatment of medulloblastoma with hedgehog pathway inhibitorGDC-0449. N Engl J Med 2009;361:1173–8.

10. Engelman JA, Janne PA. Mechanisms of acquired resistance toepidermal growth factor receptor tyrosine kinase inhibitors in non-small cell lung cancer. Clin Cancer Res 2008;14:2895–9.

11. YauchRL,DijkgraafGJ, AlickeB, Januario T, AhnCP,Holcomb T, et al.Smoothened mutation confers resistance to a Hedgehog pathwayinhibitor in medulloblastoma. Science 2009;326:572–4.

12. Pan S, Wu X, Jiang J, Gao WQ, Wan Y, Cheng D, et al. Discovery ofNVP-LDE225, a potent and selective smoothened antagonist. ACSMed Chem Lett 2010;1:130–4.

13. Buonamici S, Williams J, Morrissey M, Wang A, Guo R, Vattay A,et al. Interfering with resistance to smoothened antagonists byinhibition of the PI3K pathway in medulloblastoma. Sci Transl Med2010;2:51ra70.

14. Dijkgraaf GJ, Alicke B, Weinmann L, Januario T, West K, Modrusan Z,et al. Small molecule inhibition of GDC-0449 refractory smoothenedmutants and downstreammechanisms of drug resistance. Cancer Res2010;71:435–44.

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

16. Robinson G, Parker M, Kranenburg TA, Lu C, Chen X, Ding L, et al.Novelmutations target distinct subgroups ofmedulloblastoma.Nature2012;488:43–8.






17. Parsons DW, Li M, Zhang X, Jones S, Leary RJ, Lin JC, et al. Thegenetic landscape of the childhood cancer medulloblastoma. Science2011;331:435–9.

18. HartmannW,Digon-SontgerathB,KochA,WahaA, Endl E,Dani I, et al.Phosphatidylinositol 30-kinase/AKT signaling is activated in medullo-blastoma cell proliferation and is associated with reduced expressionof PTEN. Clin Cancer Res 2006;12:3019–27.

19. Kasper M, Jaks V, Are A, Bergstrom A, Schwager A, Svard J, et al.Wounding enhances epidermal tumorigenesis by recruiting hair folliclekeratinocytes. Proc Natl Acad Sci U S A 2011;108:4099–104.

20. Pinheiro JC, Bates DM. Mixed-effects models in S and S-PLUS. NewYork: Springer; 2000.

21. Yang ZJ, Ellis T,Markant SL, Read TA, Kessler JD, BourboulasM, et al.Medulloblastoma can be initiated by deletion of Patched in lineage-restricted progenitors or stem cells. Cancer Cell 2008;14:135–45.

22. Ellis T, Smyth I, Riley E, Graham S, Elliot K, Narang M, et al. Patched 1conditional null allele in mice. Genesis 2003;36:158–61.

23. Wong H, Alicke B, West KA, Pachero P, La H, Januario T, et al.Pharmacokinetic-pharmacodynamic analysis of vismodegib in pre-clinical models of mutational and ligand-dependent Hedgehog path-way activation. Clin Cancer Res 2011;17:4682–92.

24. Salphati L, Heffron TP, Alicke B, Nishimura M, Barck K, Carano RA,et al. Targeting the PI3K pathway in the brain–efficacy of a PI3Kinhibitor optimized to cross the blood–brain barrier. Clin Cancer Res2012;18:6239–48.

25. Kwon CH, Zhu X, Zhang J, Knoop LL, Tharp R, Smeyne RJ, et al. Ptenregulates neuronal soma size: a mouse model of Lhermitte–Duclosdisease. Nat Genet 2001;29:404–11.

26. Backman SA, Stambolic V, Suzuki A, Haight J, Elia A, Pretorius J, et al.Deletion of Pten in mouse brain causes seizures, ataxia and defects in

soma size resembling Lhermitte–Duclos disease. Nat Genet 2001;29:396–403.

27. ZhuG,ChowLM,Bayazitov IT, TongY,Gilbertson RJ, ZakharenkoSS,et al. Pten deletion causes mTorc1-dependent ectopic neuroblastdifferentiation without causing uniform migration defects. Develop-ment 2012;139:3422–31.

28. Hambardzumyan D, Becher OJ, Rosenblum MK, Pandolfi PP, Man-ova-Todorova K, Holland EC. PI3K pathway regulates survival ofcancer stem cells residing in the perivascular niche following radiationin medulloblastoma in vivo. Genes Dev 2008;22:436–48.

29. Castellino RC, Barwick BG, Schniederjan M, Buss MC, Becher O,Hambardzumyan D, et al. Heterozygosity for Pten promotes tumori-genesis in a mouse model of medulloblastoma. PLoS One 2010;5:e10849.

30. Garcia I, Crowther AJ, Gama V, Ryan Miller C, Deshmukh M,Gershon TR. Bax deficiency prolongs cerebellar neurogenesis,accelerates medulloblastoma formation and paradoxicallyincreases both malignancy and differentiation. Oncogene 2012;32:2304–14.

31. Li Y, Maitah MY, Ahmad A, Kong D, Bao B, Sarkar FH. Targeting theHedgehog signaling pathway for cancer therapy. Expert Opin TherTargets 2012;16:49–66.

32. Lin TL, Matsui W. Hedgehog pathway as a drug target: smoothenedinhibitors in development. Onco Targets Ther 2012;5:47–58.

33. Ng JM, Curran T. The Hedgehog's tale: developing strategies fortargeting cancer. Nat Rev Cancer 2011;11:493–501.

34. Lind D, Franken S, Kappler J, Jankowski J, Schilling K. Character-ization of the neuronal marker NeuN as a multiply phosphorylatedantigen with discrete subcellular localization. J Neurosci Res 2005;79:295–302.

Metcalfe et al.





2013;73:7034-7042. Published OnlineFirst October 23, 2013.Cancer Res Ciara Metcalfe, Bruno Alicke, Ailey Crow, et al. Hedgehog Pathway InhibitionPTEN Loss Mitigates the Response of Medulloblastoma to

Updated version

10.1158/0008-5472.CAN-13-1222doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/10/23/0008-5472.CAN-13-1222.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/23/7034.full#ref-list-1

This article cites 33 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/23/7034.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/73/23/7034To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-13-1222

http://cancerres.aacrjournals.org/content/suppl/2013/10/23/0008-5472.CAN-13-1222.DC1

http://cancerres.aacrjournals.org/content/73/23/7034.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/23/7034.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/23/7034


PTEN Loss Mitigates the Response of …...Medulloblastoma is the most common malignant pediat-ric...

Documents

Transcript of PTEN Loss Mitigates the Response of …...Medulloblastoma is the most common malignant pediat-ric...