Targeting Medulloblastoma: Small-Molecule Inhibitors of the Sonic

Hedgehog Pathway as Potential Cancer Therapeutics

Justyna Romer and Tom Curran

Department of Developmental Neurobiology, St. Jude Children’s Research Hospital, Memphis, Tennessee

Abstract

Medulloblastoma is the most common malignant pediatricbrain tumor for which no satisfactory treatments exist. TheSonic Hedgehog signaling pathway seems to play an impor-tant role in the pathology of this disease. Here we review ourrecent demonstration that a small-molecule inhibitor of thispathway can regress tumors that arise in a transgenic mousemodel of medulloblastoma. These and other findings suggestthat inhibitors of Sonic Hedgehog signaling may offer aneffective way to target some malignancies. (Cancer Res 2005;65(12): 4975-8)

Introduction

Current strategies to develop better cancer therapies rely to agreat extent on targets identified from studying the genetics ofhuman cancers. It is hoped that understanding the molecularpathways disrupted in cancer will yield a better understanding ofmolecular targets for therapy and thus lead to more specific andtherefore less toxic treatments. Brain tumors offer a particularlydifficult challenge not only because of their less accessibleanatomic location but also because they are relatively rare,especially in the case of pediatric brain tumors.Traditional drug development approaches for brain tumors were

optimized for cytotoxic therapies involving radiation and chemo-therapy. Although these treatments do show some efficacy, thedrawback, especially for treatment of pediatric brain tumors, arethe severe and life-long side effects (1). As yet, no specific‘‘designer’’ drug has been made for brain tumors. One of the chiefobstacles in the field is that testing new drugs relies on tissueculture and xenograft models, which recapitulate some but not allthe features of malignancy. In particular, medulloblastoma tumorcells are notoriously difficult to grow ex vivo , and due to theselective pressures imposed by cell culture, there are significantconcerns whether the cell lines used are representative of theoriginal tumor. Another consideration is that drugs are oftentested in heterogeneous populations of tumor cells that are notprofiled genetically, impeding development of targeted therapiesthat are specific for a molecular pathway and that may only bedisrupted in a subset of tumors. Some of these considerations mayexplain why previous strategies for drug development have broughtonly modest improvements to the treatment of brain tumors. Atthe current time, most advances in the clinic are still based on thestratification of patients into risk groups and optimizing the doseof radiation and chemotherapy accordingly. Therefore, thereremains a great need for more relevant models in which to

develop and test new targeted therapies. We propose here a three-step approach for preclinical development of such drugs: (a)Identify the control pathways driving tumor growth. (b) Create agenetically equivalent, high-incidence animal model, which reca-pitulates the specific subset of human disease with respect todevelopmental time frame and anatomy. (c) Design specificinhibitors of the growth control pathway impaired in tumorsand test them in the animal model. Efforts to identify new cancermodalities have increasingly relied upon advances in realizing thethree elements of this strategy.Our laboratory studies the molecular causes underlying

medulloblastoma, the most common malignant pediatric braintumor. Whereas there are several genetic subtypes of medullo-blastoma, the study described here is concerned specifically withtumors arising due to disruption of the Sonic Hedgehog (SHH)pathway. The first indication that mutations in the SHH pathwaycontribute to medulloblastoma was the finding that Gorlinsyndrome patients, who inherit germ line mutations in thePTCH1 gene, are predisposed to medulloblastoma (2, 3). PTCH1 isa trans-membrane protein that functions as a receptor for Shh. Inthe absence of Shh, PTCH1 maintains the Shh pathway in the offstate by acting on Smoothened (SMO). Upon Shh binding toPTCH1, this repression is lifted and a signal is transducedthrough Smo to the nucleus increasing expression of target genesincluding GLI1 and PTCH1 itself (Fig. 1). This pathway is crucialfor normal development of the cerebellum, because it governs theproliferation of granule neuron precursor (GNP) cells, the mostabundant neuronal population in the brain. GNP cells arebelieved to be the cells of origin for medulloblastoma, althoughdefinitive proof is still lacking. It has been suggested thatmedulloblastoma arises as an aberration of normal developmentalprocesses; a GNP cell fails to exit the cell cycle at the appropriatetime and remains in the EGL, eventually expanding to form atumor (Fig. 2; ref. 4). The finding that over 30% of sporadicmedulloblastomas show elevated expression of GLI1 supports thishypothesis (5).The contribution of an inappropriately active SHH pathway to

medulloblastoma formation makes it an ideal target to developnew therapies. In our recent study (6), we set out to investigatewhether suppressing the overactive SHH signaling in medulloblas-toma could cause tumor regression in vivo . To achieve this aim, weused an animal model of medulloblastoma (7), in which mice areheterozygous for the Ptc1 gene and null for the p53 gene. AllPtc1+/�p53�/� mice spontaneously develop medulloblastoma asearly as at 2 weeks of age and they die from the disease by week 16.The 100% incidence and short latency of tumor developmentmakes this mouse strain well-suited to validate the therapeuticpotential of Shh pathway inhibitors. The naturally occurringinhibitor of Shh signaling, cyclopamine (8, 9), which binds to Smo(10), has been shown to inhibit the growth of medulloblastomacells in cell culture and in murine allografts and xenografts (11).However, although these studies are encouraging, extrapolation of

Requests for reprints: Justyna Romer, Department of DevelopmentalNeurobiology, St. Jude Children’s Research Hospital, 332 North Lauderdale Street,Memphis, TN 38105. Phone: 901-495-4801; Fax: 901-495-2270; E-mail:stysia.romer@stjude.org.

I2005 American Association for Cancer Research.

www.aacrjournals.org 4975 Cancer Res 2005; 65: (12). June 15, 2005

Review

Research. on June 11, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

these results to appropriate tissue settings would be desired toaddress shortcomings of the artificial environments presented bytissue culture and xenograft studies. This is especially pertinent inthe case of Shh pathway–dependent tumors, as we have shownthat this pathway is down-regulated in cultured medulloblastomacells, rendering them ineffective for testing therapies based onsuppression of the Shh pathway (6).To avoid these problems we relied on in vivo studies using a

novel, small-molecule inhibitor of the Shh pathway, HhAntag-691(12), that readily enters the brain and suppresses Shh signaling bybinding to Smo (Fig. 3). This compound, a benzimidazolederivative, inhibits the Shh pathway at lower concentrations thancyclopamine. We treated Ptc1+/�p53�/� mice with different dosesof HhAntag-691, as well as with vehicle alone, and found that thetreatment led to a dose-dependent down-regulation of severalgenes that are overexpressed in medulloblastoma (Gli1 , Ptc2 , Sfrp1 ,and Math1). In contrast, Reln , a gene highly expressed inmedulloblastoma but not associated with the Shh pathway, wasnot inhibited. We also saw a reduction of Gli2 expression.1 This wasimportant, because Gli2 can to some extent compensate for loss ofGli1 in medulloblastoma formation (13). Suppression of Gli1

expression indicated that HhAntag-691 is capable of reaching itstarget in vivo , in the cerebellum of tumor-bearing mice. Furtheranalysis of tumors isolated from treated mice showed thatHhAntag-691 leads to a drastic decrease in proliferation as wellas a modest increase in tumor apoptosis. Additionally, we foundthat the treated brains became infiltrated with macrophages andreactive astrocytes, indicating that treatment incites the hostimmune response and tissue repair mechanisms. When mice weretreated for a longer period, a dramatic loss of tumor volume wasobserved. In mice treated at intermediate doses, a 7-fold loss oftumor mass was noted, and no tumor mass remained aftertreatment with the highest dose. Finally, long-term treatment withHhAntag-691 prolonged tumor-free survival of Ptc1+/�p53�/�

mice, when compared with vehicle-treated mice. Importantly, dailymonitoring of the treated mice did not reveal any weight loss orother obvious deleterious side effects. These results offer stronghope that targeting the Shh pathway in a subset of humanmedulloblastomas may provide a feasible, nontoxic therapy for thisdisease.Traditional approaches to testing new therapies, which are

based on tissue culture and murine xenograft studies, mightactually hinder development of meaningful therapies, becausethey rely on artificial systems that do not fully recapitulate thedisease in situ. Whereas it is also possible to fault some features

Figure 1. Developing targeted therapies for medulloblastoma. The Shh pathway is critical for the normal development of the cerebellum. Shh released fromthe Purkinje cells acts on overlying GNP cells in the EGL, leading to proliferation. After this period of Shh-dependent proliferation, granule neurons exit the cell cycle,begin to differentiate, and migrate inwards, past the Purkinje cell layer to reside in the IGL in the mature cerebellum.

1 Unpublished data.

Cancer Research

Cancer Res 2005; 65: (12). June 15, 2005 4976 www.aacrjournals.org

Research. on June 11, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

of transgenic mouse models of cancer, the perspective providedby the study described above illustrates a new way to developdrugs for medulloblastoma and other solid tumors that aredriven by deregulation of the Shh pathway. By directly assayingthe molecular status of the pathway being targeted, the modeloffers an invaluable tool for preclinical pharmacodynamic andpharmacokinetic experiments. With molecular targeted therapeu-tics, the traditional strategy of testing drugs at the maximumtolerated dose is clearly not appropriate. Instead, the specificnature of such agents dictates that they should be used at dosesthat can effectively suppress the target; going beyond thebiologically relevant level may lead to deleterious effects notassociated with target inhibition. Caution needs to be exercisedin applying results from mouse trials to human disease; theanimal model system described here should be used inconjunction with studies done on human tumor material, ifpossible, before moving forward to clinical trials. However, in thecase of medulloblastoma, human systems which reliably recapit-ulate the in situ disease are not yet available. If the mouse modelrecapitulates the genetic signature of the human disease, as wellas its anatomic and temporal characteristics, targeting the samepathway in human disease may also be effective. One issue formedulloblastoma is that it is predominantly a pediatric tumor,and new treatments for cancer are usually tested first in adult

patients. For this reason, the hope for clinical trials will likelystart with tests on adult tumors that rely on Shh signaling,thereby paving the way for clinical trials for medulloblastoma inchildren. Basal cell carcinoma (BCC) is one example of a tumorin which the initial genetic lesion (PTCH1 mutation) is the sameas in medulloblastoma, making BCC suitable for evaluating theefficacy of Shh suppression as a therapy (14). Additionally, recentreports point to other solid tumors that are dependent on Shhsignaling for growth, although genetic lesion in the Shh pathwayhave yet to be identified in these cases (15–17). Such tumorsoffer additional candidates for initial clinical trials of Shhinhibitors. We hope that the efforts to produce animal modelswhich recapitulate specific human diseases will offer animproved chance of developing nontoxic, targeted cancertherapies, each one designed for a specific subtype of tumor,based on its genetic profile. This would be especially beneficial inthe area of pediatric brain tumors, where the currently availablecytotoxic treatments leave patients to cope with sometimes verysevere, neurologic and neuroendocrine consequences for the restof their lives.

Acknowledgments

Received 2/11/2005; revised 4/5/2005; accepted 4/22/2005.

Figure 2. Medulloblastoma can arise as an aberration of cerebellar development, when a GNP cell fails to exit the cell cycle and remains proliferating in the EGL.Somatic mutations in different components of Shh pathway, including PTCH1 and SMOH may contribute to sporadic medulloblastoma (18–21) and germ linemutations of PTCH1 gene underlie tumors arising in patients with Gorlin syndrome (22, 23).

Targeting Medulloblastoma

www.aacrjournals.org 4977 Cancer Res 2005; 65: (12). June 15, 2005

Research. on June 11, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

References1. Gilbertson RJ. Medulloblastoma: signalling a change intreatment. Lancet Oncol 2004;5:209–18.

2. Goodrich LV, Scott MP. Hedgehog and patched inneural development and disease. Neuron 1998;21:1243–57.

3. Kimonis VE, Goldstein AM, Pastakia B, et al. Clinicalmanifestations in 105 persons with nevoid basal cellcarcinoma syndrome. Am J Med Genet 1997;69:299–308.

4. Wechsler-Reya R, Scott MP. The developmental biologyof brain tumors. Annu Rev Neurosci 2001;24:385–428.

5. Lee Y, Miller HL, Jensen P, et al. A molecular fingerprintfor medulloblastoma. Cancer Res 2003;63:5428–37.

6. Romer JT, Kimura H, Magdaleno S, et al. Suppressionof the Shh pathway using a small molecule inhibitoreliminates medulloblastoma in Ptc1(+/�)p53(�/�)mice. Cancer Cell 2004;6:229–40.

7. Wetmore C, Eberhart DE, Curran T. Loss of p53 butnot ADP ribosylation factor accelerates medulloblasto-ma in mice heterozygous for patched. Cancer Res 2001;61:513–6.

8. Binns W, James LF, Shupe JL, Everett G. A congenitalcyclopian-type malformation in lambs induced bymaternal ingestion of a range plant, Veratrum californi-cum . Am J Vet Res 1963;24:1164–75.

9. Keeler RF, Binns W. Teratogenic compounds ofVeratrum californicum (Durand). V. Comparison of

cyclopian effects of steroidal alkaloids from the plantand structurally related compounds from other sources.Teratology 1968;1:5–10.

10. Chen JK, Taipale J, Cooper MK, Beachy PA. Inhibitionof hedgehog signaling by direct binding of cyclopamineto smoothened. Genes Dev 2002;16:2743–8.

11. Berman DM, Karhadkar SS, Hallahan AR, et al.Medulloblastoma growth inhibition by hedgehog path-way blockade. Science 2002;297:1559–61.

12. Gabay L, Lowell S, Rubin LL, Anderson DJ. Deregu-lation of dorsoventral patterning by FGF conferstrilineage differentiation capacity on CNS stem cellsin vitro . Neuron 2003;40:485–99.

13. Kimura H, Stephen D, Joyner A, Curran T. Gli1 isimportant for medulloblastoma formation in Ptc1(+/�)mice. Oncogene 2005.

14. Williams JA, Guicherit OM, Zaharian BI, et al.Identification of a small molecule inhibitor of thehedgehog signaling pathway: effects on basal cellcarcinoma-like lesions. Proc Natl Acad Sci U S A 2003;100:4616–21.

15. Berman DM, Karhadkar SS, Maitra A, et al.Widespread requirement for hedgehog ligand stimula-tion in growth of digestive tract tumours. Nature 2003;425:846–51.

16. Thayer SP, di Magliano MP, Heiser PW, et al.Hedgehog is an early and late mediator of pancreaticcancer tumorigenesis. Nature 2003;425:851–6.

17. Watkins DN, Berman DM, Burkholder SG, Wang B,Beachy PA, Baylin SB. Hedgehog signalling within airwayepithelial progenitors and in small-cell lung cancer.Nature 2003;422:313–7.

18. Raffel C, Jenkins RB, Frederick L, et al. Sporadicmedulloblastomas contain PTCH mutations. Cancer Res1997;57:842–5.

19. Pietsch T, Waha A, Koch A, et al. Medulloblastomasof the desmoplastic variant carry mutations of thehuman homologue of Drosophila patched. Cancer Res1997;57:2085–8.

20. Vorechovsky I, Tingby O, Hartman M, et al. Somaticmutations in the human homologue of Drosophilapatched in primitive neuroectodermal tumours. Onco-gene 1997;15:361–6.

21. Wolter M, Reifenberger J, Sommer C, Ruzicka T,Reifenberger G. Mutations in the human homologue ofthe Drosophila segment polarity gene patched (PTCH)in sporadic basal cell carcinomas of the skin andprimitive neuroectodermal tumors of the centralnervous system. Cancer Res 1997;57:2581–5.

22. Hahn H, Wicking C, Zaphiropoulous PG, et al.Mutations of the human homolog of Drosophila patchedin the nevoid basal cell carcinoma syndrome. Cell 1996;85:841–51.

23. Johnson RL, Rothman AL, Xie J, et al. Humanhomolog of patched, a candidate gene for the basal cellnevus syndrome. Science 1996;272:1668–71.

Figure 3. The link between the Shh pathway and medulloblastoma led us to investigate whether suppressing Shh signaling in these tumors in vivo could betherapeutic. We used a small molecule inhibitor of the Shh pathway (HhAntag-691, a Smo inhibitor) in a mouse model of medulloblastoma. We found that treating micewith the Smo inhibitor suppressed the Shh pathway and led to the elimination of tumors. This was accompanied by a decrease in tumor cell proliferation and anincrease in apoptosis. Importantly, long-term treatment prolonged tumor-free survival of the mice. This experiment underlines the importance of developingrelevant animal models in which to test new cancer therapies.

Cancer Research

Cancer Res 2005; 65: (12). June 15, 2005 4978 www.aacrjournals.org

Research. on June 11, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

2005;65:4975-4978. Cancer Res   Justyna Romer and Tom Curran  Sonic Hedgehog Pathway as Potential Cancer TherapeuticsTargeting Medulloblastoma: Small-Molecule Inhibitors of the

  Updated version

  http://cancerres.aacrjournals.org/content/65/12/4975

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/65/12/4975.full#ref-list-1

This article cites 22 articles, 9 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/65/12/4975.full#related-urls

This article has been cited by 25 HighWire-hosted articles. Access the articles at:

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/65/12/4975To request permission to re-use all or part of this article, use this link

Research. on June 11, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from