The Normal patchedAllele Is Expressed in Medulloblastomas from Mice withHeterozygous Germ-Line Mutation of patched1
Cynthia Wetmore, Derek E. Eberhart, and Tom Curran2
Departments of Developmental Neurobiology [C. W., D. E. E., T. C.] and Hematology/Oncology [C. W.], St. Jude Children’s Research Hospital, 332 North Lauderdale, Memphis,TN 38105
ABSTRACT
Defects in a developmental signaling pathway involving mammalianhomologues of theDrosophilasegment polarity gene,patched(ptc) and itsligand, sonic hedgehog(shh), contribute to tumor formation in severaltissues. Recently, a subset of medulloblastoma, the most common malig-nant brain tumor in children, was found to contain somatic mutations inthe human ptc gene. In addition, basal cell nevus syndrome (BCNS), orGorlin syndrome, which is characterized by developmental anomalies anda predisposition to skin and nervous system malignancies, is associatedwith germ-line mutation of ptc. Targeted disruption of both alleles ofptcin mice results in embryonic lethality. However,ptc1/2 mice survive anddevelop spontaneous cerebellar brain tumors, suggesting thatptc mayfunction as a tumor suppressor gene. Therefore, we investigatedptc1/2
mice as a model for human medulloblastoma. We report that 14% ofptc1/2 mice develop central nervous system tumors in the posterior fossaby 10 months of age, with peak tumor incidence occurring between 16 and24 weeks of age. The tumors exhibited several characteristics of humanmedulloblastoma, including expression of intermediate filament proteinsspecific for neurons and glia. Full-length ptc mRNA was present in alltumors analyzed, indicating that there was no loss of heterozygosity at theptc locus. Nucleotide sequence ofptc mRNA from four tumors failed toidentify any mutations. However, a comparison of the normalptcsequencefrom C57BL/6 and 129Sv mice did reveal several polymorphisms. Highlevels ofgli1 mRNA and protein were detected in the tumors, suggestingthat the shh/ptc pathway was activated despite the persistence ofptcexpression. These data indicate that haploinsufficiency ofptc is sufficientto promote oncogenesis in the central nervous system.
INTRODUCTION
Medulloblastoma, a primitive neuroectodermal tumor that arises inthe cerebellum, is the most common malignant brain tumor in chil-dren. Peak incidence occurs at 5 years of age, and most tumors appearduring the first decade of life (1). Although comparative genomehybridization identified several candidate chromosomal rearrange-ments in medulloblastoma, no specific fusion gene products wereidentified (2, 3). However, a subset of medulloblastoma has beenidentified with allelic loss of chromosome 9q22 (4–8), a region thatcontainsptc3 (9, 10). In contrast to numerous reports of p53 mutationin adult brain tumors, mutations in p53 have only rarely been dem-onstrated in pediatric medulloblastoma (11). The discovery that thehuman homologue of theDrosophila segment polarity gene,ptc, ismutated in a subset of spontaneous (4–8) and basal cell nevussyndrome-associated medulloblastoma (12) provided the first genetic
clue to the etiology of medulloblastoma.ptc mutations have beendetected in both desmoplastic and classic medulloblastoma, and theycannot be correlated with a specific tumor subtype (7).
Theptc gene was identified during a genetic screen for embryoniclethal mutations inDrosophila. Alterations inptc are associated withaberrant body segmentation of the larva and early lethality (13).ptcencodes a 12-pass transmembrane protein (14) that functions as partof the receptor complex forhh (15). During development,ptc re-presses constitutive signaling bysmo, another component of thehhreceptor complex. This results in reduced transcription of severaltarget genes in the pathway including members of thetransforminggrowth factorb family, wnt genes,ci, andptc itself (16). Binding ofhh to theptc/smoreceptor complex relieves this ptc-mediated repres-sion and results in increased activity of the transcription factorci,which elevates expression of target genes (17).
In vertebrates, mutations in several members of thehh/ptcpathwayhave been linked to developmental defects in the nervous system andto malignancies in several tissues. Mutations inshh, a vertebratehomologue ofhh, are associated with neural tube defects in humans,including spina bifida and holoprosencephaly (18, 19). In humantumors, inactivation ofptc abrogates its repressor function, increasingthe level of expression ofgli1, a vertebrate transcription factor relatedto ci, that was originally identified as an amplified gene in a glioblas-toma (20). Although it is present in glioblastomas and sarcomas,gli1is not readily detected in adult brain RNA (21). Overexpression ofgli1in the epidermis of frogs causes lesions resembling human BCC,similar to those induced in mice by overexpression ofshh (19). Thehh/ptc pathway is thought to play a critical role in regulating cellproliferation and differentiation during morphogenesis.
Germ-line mutations inptc are responsible for BCNS (also knownas Gorlin Syndrome; OMIM 109400), an autosomal dominant disease(22) characterized by a larger body size, developmental and skeletalanomalies, fibromas of soft tissues, radiation sensitivity, basal cellcarcinoma, and medulloblastoma (19, 23). Although only a smallproportion of medulloblastoma is associated with BCNS (10% ofchildren diagnosed with medulloblastoma at age 2 years or under; Ref.12), these patients are very sensitive to radiation, which is often usedin the treatment of medulloblastoma. In BCNS patients,ptcappears tofunction as a classic tumor suppressor gene because the second alleleis lost in the majority of BCCs (12). Mice heterozygous forptcresemble BCNS patients in that they exhibit a larger body size,skeletal abnormalities, cerebellar tumors (24), radiation sensitivity(25), and skin lesions similar to BCC after radiation (26). In mice,homozygous loss ofptc results in embryonic lethality at 9.5–10.5 daysafter fertilization (24).
We investigated tumorigenesis inptc1/2 mice to assess the utilityof this strain as a model for pediatric medulloblastoma. In our colony,we found that 14% ofptc1/2 mice develop posterior fossa tumors by10 months of age. The histological appearance, site of origin, andpresence of intermediate filament proteins indicated that the mousetumors were very similar to human medulloblastoma. In contrast tothe situation with BCC in mice and humans, we found that the normalptc allele was retained and expressed in mouse medulloblastomas.Thus, haploinsufficiency ofptc promotes medulloblastoma formation
Received 12/22/99; accepted 3/2/00.The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.
1 Supported in part by NIH Cancer Center Support CORE Grant P30 CA 21765, theAmerican Lebanese Syrian Associated Charities, the Pediatric Bear Necessities ResearchFoundation (to C. W.), the Pediatric Brain Tumor Foundation of the United States (toC. W.), NIH training grant for Physician-Scientists (to C. W.), and an American CancerSociety Postdoctoral Fellowship (to D. E.).
2 To whom requests for reprints should be addressed, at Department of DevelopmentalNeurobiology, St. Jude Children’s Research Hospital, 332 North Lauderdale, Memphis,TN 38105. Phone: (901) 495-2255; Fax: (901) 495-2270; E-mail: [email protected].
in mice by a mechanism that does not require complete loss ofptcexpression.
MATERIALS AND METHODS
Animals. A colony of ptc1/2 mice was established using two pairs of mice,obtained from Matthew Scott, Stanford University. These mice were either crossedwith C57BL/6 or 129Sv (currently designated by The Jackson Laboratory as1293 1/SvJ) mice. Mice were maintained in the animal facility at SJCRH on a12-h light/dark cycle. Animals were genotyped by PCR amplification (24) ofgenomic DNA extracted from tail biopsies. Approximately 1000 mice (200 129Svand 800 C57BL/6) were generated and genotyped to obtain 409ptc1/2 mice forthe present study. Heterozygous mice were maintained and observed up to 10months of age. Animals were sacrificed when they showed signs of increasedintracranial pressure (ataxia, decreased movement, paresis of hind limbs, enlargedoccipital prominence, hunched back, and/or poor grooming)
Dissection of Brains and Tumor Tissues.Brains were removed from thesurrounding calvarium, and tumor tissue was carefully separated from sur-rounding brain parenchyma under a dissecting microscope. Generally, therewas no capsule surrounding tumors, and malignant tissue was separated fromuninvolved brain by blunt dissection based on the difference in tissue consis-tency and the anatomy of the cerebellum. Fresh tissue was snap frozen andstored at280°C for later extraction of RNA, DNA, and protein. For histo-chemical analyses, animals were deeply anesthetized and perfused transcardi-ally with 4% paraformaldehyde in 0.1M PBS.
Histochemistry. After perfusion, tissues were postfixed in 4% paraformalde-hyde in 0.1M PBS (pH 7.4) for 2–6 h at room temperature. Subsequently, theywere immersed in 20% sucrose for cryoprotection and stored at 4°C. Tissue wasembedded in freezing medium, frozen, sectioned at a thickness of 12mm on acryostat, and mounted onto Superfrost1 slides (Fisher Scientific, Pittsburgh, PA).
Frozen sections of tumor tissue were treated with antibodies to neurofila-ment-H (NF; 1:200; Chemicon International, Temecula, CA), GFAP (1:200;Dako A/S, Copenhagen, Denmark), vimentin (1:200; Zymed Laboratories, SanFrancisco, CA), Nestin (1:50; Rat401; Iowa Hybridoma Tissue Bank), andsynaptophysin (1:200; Chemicon International). Primary antibodies were di-luted in 0.1M PBS with 0.1% Triton X-100 and applied to tissue overnight at4°C. Secondary antibodies were applied according to Vectastain Elite ABCinstructions (Vector Laboratories, Burlingame, CA), and detection was carriedout with 3,39-diaminobenzidine reagent (Kirkegaard & Perry Laboratories,Gaithersburg, MD).
RNA Isolation and Northern Analysis. Total cellular RNA was isolatedfrom 18 mouse medulloblastomas using Trizol (Ambion, Inc., Austin, TX)according to the manufacturer’s directions. Five to 10mg of total RNA wereelectrophoresed on a 0.8% agarose-formaldehyde gel, transferred to a nitro-cellulose filter (Hybond N1; Amersham Pharmacia, Buckinghamshire, Eng-land), and hybridized under stringent conditions (18 h at 68°C in 53SSPE,50% formamide, 53Denhardt’s solution, 1% SDS, and 0.1 mg/ml denaturedsalmon sperm DNA) with a32P-labeled RNA probe. Filters were washed(twice for 20 min each in 0.13SSC, 0.1% SDS at 68°C) and exposed to MRfilm (Kodak) for 12–72 h at280°C. Control andptc1/2 tissues were analyzedby hybridization with32P-labeled RNA probes specific for mouseptc (24),gli1(Mouse EST clone 38654),gli3 (27), reelin (28), trkC (29), and random-primed DNA probes specific forlacZ and b-actin (RediPrime; AmershamPharmacia Biotech UK).
Generation of Gli1 Antisera. Antibodies were generated in rabbits againstCOOH region peptides corresponding to amino acids 803–818 of mouse Gli1.Peptides were synthesized as multiple antigen peptides (Research Genetics,Huntsville, AL) to avoid generating antibodies to a carrier protein. Rabbitswere immunized with 1 mg of peptide in Freund’s Complete adjuvant andboosted approximately every 2 weeks with 0.5–0.25 mg of peptide in incom-plete adjuvant, and blood was collected via the central ear artery at 10 daysafter immunization. Crude sera were collected after allowing the blood to clotovernight at 4°C and used at stated dilutions without further purification. Serawere screened in immunoblot, immunofluorescence, and immunoprecipitationassays and selected for further evaluation.
Transfection and Immunoblot Analysis. COS-7 cells or 293T cells weretransfected with an expression plasmid encoding humangli1 cDNA (30) usingFuGene6 (Boehringer Mannheim) according to the manufacturer’s directions.
Forty-eight h after transfection, the cells were lysed in 23SDS sample buffer(100 ml per 35-mm dish) for immunoblot analysis. As a positive control fornative Gli1, we analyzed protein extracts from the rhabdomyosarcoma cell lineRh30, which expresses high levels ofgli1 (31).
Protein extracts were prepared by Dounce homogenization of 80–100 mg ofsnap-frozen tumor or normal tissue in 300ml of ice-cold lysis buffer [1%Triton X-100, 30 mM HEPES (pH 7.5), 10% glycerin, 150 mM NaCl, 5 mM
MgCl2, 1 mM EGTA, 0.1%b-mercaptoethanol, 2 mM phenylmethylsulfonylfluoride, 10mg/ml aprotinin, 10mg/ml leupeptin, 10mg/ml trypsin inhibitor,and 1mg/ml pepstatin]. Extracts were cleared by microcentrifugation at 14,000rpm for 30 min. One hundredmg of the protein extracts or 6ml of 293T celllysates were loaded per lane on an 8% polyacrylamide gel. After separation ona polyacrylamide gel, proteins were electrotransferred onto nitrocellulosemembranes, incubated with an antibody directed against mouse Gli1 aminoacids 803–818 (1:5000), followed by incubation with a rabbit IgG conjugatedto horseradish peroxidase at 40 milliunits/ml, and products were visualized byenhanced chemiluminescence (Boehringer Mannheim). Membranes werestripped and incubated with an antibody against Ref-1 (32) to control forprotein loading and transfer efficiency.
Reverse Transcription-PCR. Two-step reverse transcription-PCRs werecarried out to maximize uniformity of PCR templates for all reactions. cDNA wasderived in 20ml volumes with random hexamers, oligo dT, and gene-specificpriming using SuperScript reverse transcriptase (Life Technologies, Inc., Rock-ville, MD). Reverse transcriptase first-strand cDNA synthesis reactions werecarried out using 3mg of total RNA prepared from adult C57BL/6 and 129Svmouse cerebellum, C57BL/6 E15 limbs, and four tumor samples (tumor numbers199, 241, 448, 646), according to the manufacturer’s directions. PCR productswere analyzed from two separate cDNA synthesis reactions.
Nucleotide Sequencing.Sequencing reactions were performed by theHartwell Center for Biotechnology at St. Jude Children’s Research Hospital ontemplate DNA using rhodamine or dRhodamine dye-terminator cycle sequenc-ing ready reaction kits with AmpliTaq DNA polymerase FS (Perkin-Elmer,Applied Biosystems, Inc., Foster City, CA), and synthetic oligonucleotides.Samples were electrophoresed, detected, and analyzed on Perkin-Elmer Ap-plied Biosystems model 373, model 37 Stretch, or model 377 DNA sequencers.Sequence analysis was performed using Sequencher (Gene Codes Corp., AnnArbor, MI) software.
Restriction Enzyme Digest of PCR Products.The nucleotide polymor-phism at position 4015 ofptc in C57BL/6 and 129Sv mouse DNA disrupts therecognition site for the restriction enzyme,MaeII (A/CGT; Boehringer Mann-heim). To differentiate between the polymorphic alleles, PCR fragments am-plified from tail DNA from 129Sv mice, C57BL/6 mice, and tail DNA fromfour ptc1/2 mice (nos. 199, 241, 448, and 646) that developed tumors were
Fig. 1. Histogram of medulloblastoma incidence inptc1/2 mice. All brain tumors arosein the posterior fossa and were detected by physical examination of the mice for signs ofincreased intracranial pressure. Brains were dissected and examined to confirm thepresence of tumors. Mouse age at time of tumor detection was calculated to the nearestweek at time of death.
treated withMaeII. The PCR reaction generated a 524-bp fragment from the 39end of theptc open reading frame (primersptcF8 corresponding to bases3781–3804 andptcR8 complementary to bases 4283–4305).MaeII digestionof 0.5 mg of each PCR product was performed with buffer supplied bymanufacturer. The presence of the C57BL/6 and 129Sv alleles was determinedby comparison of untreated (2) andMaeII-treated PCR products by agarosegel electrophoresis in the presence of ethidium bromide.
RESULTS
Mice Heterozygous forptc Develop Brain Tumors in the Pos-terior Fossa. Fourteen % ofptc1/2 mice between the ages of 8 and48 weeks of age developed spontaneous intracranial tumors resem-bling medulloblastoma that became large enough to cause a change innormal behavior. There was no significant difference in tumor inci-
Fig. 2. Gross and histological appearance ofmedulloblastoma inptc1/2 mice.A, gross appearanceof normal adult C57BL/6 mouse brain at 4 months ofage;B, Cresyl violet-stained sagittal section of adultC57BL/6 mouse brain;C, gross appearance of tumorin the posterior fossa ofptc1/2 mouse brain at 4months;D, Cresyl violet-stained sagittal section of thesameptc1/2 mouse brain shown inC. theboxed areais shown at higher magnification inE; E, Cresylviolet-stained section of area outlined in Fig. 5Dthrough theptc1/2 tumor. Note that tumor cells havescant cytoplasm, and they compress and displace nor-mal brain tissue. There was no infiltration of tumorcells into white matter tracts.F, GFAP immunoreac-tivity showing filamentous pattern of localization inpatches throughout the tumor.G, NF immunoreactiv-ity in cytoplasm of tumor cells showing filamentouspattern of localization in patches throughout the tu-mor.H, synaptophysin immunoreactivity in cytoplasmof a majority of tumor cells.Bar, 30 mm for E–H.
dence between the two mouse strains (C57BL/6, 14%; 129Sv, 12%).The earliest tumors were detected at 8–10 weeks of age (Fig. 1). Peakincidence occurred between 16 and 24 weeks, although some tumorsarose in mice older than 28 weeks of age (Fig. 1). The tumors werereadily apparent on gross examination of the brains (Fig. 2C), andthey arose consistently in the posterior fossa, generally displacingnormal cerebellar tissue (Fig. 2D). The surface of the tumor wassmooth (Fig. 2,C andD) and lacked cerebellar foliations (Fig. 2,AandB). Five % of theptc1/2 mice developed an enlarged cranial vaultwith a thin membrane bridging the area in which calvarial sutureswould form in normal mice. These mice had significantly compressedcortical layers with enlarged ventricles, consistent with hydrocephalus(Table 1). Four % of the mice exhibited skeletal anomalies (polydac-tyly) or soft tissue tumors arising in skeletal muscle and connectivetissues (Table 1). Brains from;30 asymptomatic adult heterozygousmice were examined grossly under a32.5 dissecting microscope. Notumors were detected, although small regions of focal thickening ofthe cerebellar folia were noted in three mice.
Histological Analysis of Tumors. Histological analysis revealedthat the tumors, which lacked a distinct capsule, were comprised ofdensely packed cells with prominent nuclei and scant cytoplasm. Inthe majority of cases, the tumors displaced and compressed the normalcerebellar architecture (Fig. 2E). Tumor cells did not infiltrate normalareas of the cerebellum, and they did not migrate along white mattertracts. The tumors were well circumscribed by a border of compressedcells that expressed intermediate filament proteins of neuronal line-age, including NF (Fig. 2G) and synaptophysin (Fig. 2H). GFAPimmunoreactivity was noted in patches of cells throughout the tumormass (Fig. 2F). It was not possible to determine whether a single cellexpressed both NF and GFAP, although there were significant areas ofoverlapping GFAP and NF immunoreactivity. A few scattered areasof vimentin immunoreactivity were detected in three of the fivetumors analyzed. No grossly apparent lesions were noted elsewhere inthe neuroaxis. On the basis of the histological appearance, location,and expression of markers, the tumors inptc1/2 mice closely resem-ble human medulloblastoma.
Persistence of Normalptc mRNA in Tumors. Northern analysisof mRNA extracted from medulloblastomas arising inptc1/2 miceindicated that the normalptc allele was expressed in all tumorsexamined (Fig. 3). Two majorptc transcripts were detected in bothtumor and control tissues that migrated slower than 28S rRNA, atapproximately 8 and 12.5 kb (Fig. 3). In 5 of the 10 tumor samples,a smaller transcript (;5 kb) was also detected (tumors of mice nos.42, 185, 199, 530, and 646) and in E15 limb. The levels ofptcmRNAdetected in tumors varied, but it was always less than that observed inthe adult mouse cerebellum (Fig. 3). Northern analysis was alsocarried out using a lacZ probe to detect mRNAs derived from thetargeted allele. The lacZ probe identified a distinct but less abundantRNA species in tumor samples that was larger than the normalptcmRNAs and was not present in RNA extracted from normal tissues.
gli1 Is Expressed in Tumors.Transcription ofgli1, a member ofthe gli family of transcription factors, is normally repressed byptcsignaling (19).gli1 expression is increased after the interaction of Shhwith the Ptc/Smo receptor complex, which relieves the repression
function of Ptc, or after inactivation ofptcby mutation. Therefore, wecompared the expression ofgli1 in ptc1/2 brain tumors and in controltissues to determine whether theptc pathway was activated. A singlegli1 mRNA transcript of;4 kb was present at low levels in normalandptc1/2 brain tissues; however, this mRNA species was present atmuch higher levels in all tumors examined (Fig. 3). The increasedexpression ofgli1 mRNA observed in tumors compared with adultcontrol tissues implies thatgli1 transcription is derepressed inptc1/2
tumors. In contrast, there was no expression ofgli3, anothergli familymember, in the tumor tissues. However,gli3 mRNA expression wasdetected in E15 and E19 limbs.
We generated antisera to a region of mouse Gli1 (amino acids803–818) that is not conserved in Gli2 and Gli3. Protein extracts wereprepared from the group of 10 tumors analyzed in Fig. 3 and examinedusing anti-Gli1 antiserum. A single major protein ofMr ;150,000 wasidentified in tumor extracts, in the Rh30 rhabdomyosarcoma cellwhich expresses high levels ofgli1, in 293-T cells transfected withhumangli1 cDNA, in ptc1/2 brain tumor tissues, and, at lower levels,in normal E15 limb (Fig. 4). However, Gli1 protein was not detectedin normal brain, cerebellum, nor in ptc1/2 cerebellum. The high levelof expression ofgli1 mRNA and protein in tumors suggests that theshh/ptcpathway is activated in these tumors despite the presence ofptc mRNA.
Nucleotide Sequence Analysis ofptc mRNA and Detection ofPolymorphisms. The secondptc allele is mutated frequently in skintumors from individuals with a heterozygous germ-line mutation inptc (9, 10). To determine whether the normal allele ofptcwas mutatedin medulloblastomas fromptc1/2 mice, we compared the nucleotidesequence ofptc from four tumors, C57BL/6 cerebellum and C57BL/6
Fig. 3. Northern blot analysis of total RNA prepared from tumors and control tissues. Tenmg of total RNA were loaded per lane from each of the following samples: postnatal day 7C57BL/6 mouse brain (P7 brain), C57BL/6 adult cerebellum (1/1 CB), nontumor containingadultptc1/2 cerebellum (1/2 CB), tumors from 10ptc1/2 mice that developed tumors in theposterior fossa (designated by mouse nos. 42, 169, 185, 199, 241, 448, 530, 574, 646, and659), limbs from embryonic day 15 (E15 limb) and embryonic day 19 (E19 limb) C57BL/6mouse. Total RNA was hybridized with [32P]dCTP-labeled gene-specific probes forptc, gli1,reelin, trkC,andactin as designated and exposed to X-ray film for detection. Two transcripts(8 and 12.5 kb) were detected in all samples with theptcprobe, as well as a smaller transcript(;5 kb) detected in 5 of 10 tumors (nos. 42, 185, 199, 530, and 646). Some degradation ofmRNAs in samples 241 and 574 have likely occurred.
Table 1 Incidence of tumors and developmental anomalies in ptc1/2 mice
embryonic limbs. To ensure sequence fidelity, we carried out twocomplete sequencing passes of theptc open reading frame using twodifferent PCR reactions for both sense and antisense strands ofptccDNA. Each PCR reaction was repeated with two independent prep-arations of cDNA from each sample. No mutations were detected inany of the four tumors examined. However, several sequence discrep-ancies were uncovered between the tumors and the normalptc se-quence from C57BL/6. Sequence analysis of selected regions of tailDNA from the two mouse strains, C57BL/6 and 129Sv, used in thisstudy demonstrated that the sequence discrepancies corresponded tofive polymorphisms (Table 2 and Figs. 5 and 6). One of the fivepolymorphisms (C or T at nucleic acid residue 4015) resulted in acodon change from ACG (threonine in C57BL/6 mice) to ATG(methionine in 129SvJ mice) at amino acid residue 1339 (Table 2). Infact, both of these alternatives were detected in the sequence chro-matogram of the PCR products obtained from genomic DNA ex-tracted fromptc1/2 mouse 448. The presence of both alleles gener-ated two peaks on the chromatogram at position 4015 (Fig. 5). Thispolymorphism generates aMaeII restriction site in the C57BL/6ptcsequence that is not present in the 129Svptc sequence (Fig. 6).Treatment of the PCR product amplified from the C57Bl/6 allele (butnot the 129Sv allele) resulted in two fragments of 235 and 289 bases
that migrated as a close doublet on the 1% agarose gel (Fig. 6). Thedisruptedptc allele, which would be expected to be present in allheterozygous mice, is derived from 129Sv ES cells; therefore, it is notcleaved at this site byMaeII. Mice crossed onto the 129Sv back-ground (e.g.,mouse 241) contain a normal 129Svptc allele as well asthe targeted 129Sv allele, neither of which is digested byMaeII (Fig.6). In contrast, the three otherptc1/2 mice examined, which werecrossed with C57BL/6 mice (i.e., mice nos. 199, 448, and 646), havea normalptc sequence derived from C57BL/6 containing an intactMaeII restriction site (Fig. 6). Because these are heterozygous mice,they should also carry the targeted 129Sv allele, which does not
Fig. 4. Immunoblot analysis of Gli1 protein expression in mouse tumors. Protein lysates from tumors (mouse nos. 42, 169, 185, 199, 241, 448, 530, 574, 646, and 659), normalcerebellum, normal brain excluding cerebellum,ptc1/2 cerebellum, normal mouse embryonic day 15 limb (E15 Limb), E15 head, 293 cells transfected with vector alone or with hGli1,and the rhabdomyosarcoma tumor cell line Rh30, which is known to express high levels ofgli1 mRNA (31), were separated on 8% polyacrylamide gels and transferred to nitrocellulosemembranes. The membranes were incubated with anti-Gli1 serum, followed by antirabbit IgG-horseradish peroxidase, and the signal was detected by enhanced chemiluminescence.The membranes were stripped and incubated with an antibody directed against Ref-1 as a control for protein loading and transfer efficiency.
Fig. 5. Chromatogram of nucleic acid sequence analysis of a PCR fragment amplifiedfrom tail DNA extracted from mouse 448. The PCR reaction amplified regions from bothC57BL/6- and 129Sv-derivedptc alleles, creating a mixed template for sequencing. Notethat there are two peaks at position 4015 representing the mixed template, indicating bothA and G (corresponding to T and C in the sense strand ofptc).Table 2 Polymorphismsa in ptc sequence from C57BL/6 and 129Sv mice
BaseC57BL/6J
codon129SVcodon Result in amino acid sequence
3180 (T or C) ATT ATC No change3318b GAG GAG No change3438 (C or T) ACC ACT No change3498 (G or A) CCG CCA No change3561 (T or C) CCT CCC No change4015 (C or T) ACG ATG Thr or Met at 1339a Polymorphisms inptc were identified by nucleotide sequence analysisptc from
C57BL/6 and 129Sv mouse tail DNA. Discrepancies in nucleic acid sequence between thetwo strains and comparison of our sequencing with the mouseptc sequence in GenBankare highlighted in bold typeface in the codon columns.
b One nucleic acid discrepancy was identified at nucleotide 3318 between the mouseptcsequence registered in GenBank (accession no. U46155) and theptcsequence obtainedfrom both mouse strains in the present study. This discrepancy does not result in an aminoacid change.
contain aMaeII site. Therefore, this restriction enzyme can be used todetermine the presence or absence of C57BL/6 or 129Sv alleles intumor DNA. Surprisingly, one of the tumors analyzed arose in aptc1/2 mouse (no. 646) that did not retain the 129Sv allele (Fig. 6).However, this mouse did carry the 59region of the targetedptc allele,as evidenced by PCR amplification of neo-containing sequences fromgenomic tail DNA. Thus, the disruptedptc allele in mouse 646 mayhave undergone an additional deletion or rearrangement.
Expression ofreelin mRNA. To further characterize the potentialcell of origin of the medulloblastomas in ptc1/2 mice, we investigatedwhether tumor cells expressed mRNA forreelin, an extracellularprotein that directs cell migration in the developing brain (28).reelinis expressed at high levels in granule cell precursors in the cerebellumduring early stages of postnatal development (28). A singlereelinmRNA transcript of;12 kb was detected in all tumors examined (Fig.3). The presence of a distinct band of high molecular weight speciesof reelinmRNA confirmed the integrity of the tumor-derived mRNA andprovided evidence that the tumors may have arisen from granule cellprecursors rather than from a cell of glial lineage in the cerebellum.
Expression of trkC in Tumors. trkC encodes a receptor for theneurotrophic factor neurotrophin-3, which is widely expressed bygranule cells in the cerebellum during development (29).trkC expres-sion by medulloblastoma cells has been correlated with a subset oftumors that have a more favorable prognosis for event-free survival(33). Therefore, we examinedptc1/2 tumor RNA for the presence oftrkC transcripts. TwotrkC mRNA species, of approximately 6 and 4.7kb, were present at elevated levels in tumors compared with devel-oping brain or adult cerebellum (Fig. 3). There was no apparentcorrelation among levels of expression oftrkC, ptc,or gli1 mRNA inthe tumors examined.
DISCUSSION
During formation of the nervous system, the proliferation, differ-entiation, and migration of many cell types are tightly regulated bycell-cell interactions, the production of morphogens and growth fac-tors, and the selective expression of receptors for these molecules(34). The signaling pathway involvingshh,ptc,smo,and several othergenes plays a critical role in development of the cerebellum. Shh,released by Purkinje cells, relievesptc-mediated repression of cellgrowth and promotes proliferation of early granule cell precursors (35,36). Granule cells represent the most abundant cell type in the brain(they comprise more than half of the neurons in the adult mouse), andthey undergo dramatic expansion by proliferation of precursors in theexternal germinal layer of the cerebellum during development. Afterthis proliferative phase, granule cells exit the cell cycle and migrate
inwardly along radial fibers to assume their mature position in theinternal granule layer (37). Medulloblastoma is thought to arise froma primitive neuroectodermal cell, perhaps a precursor of cerebellargranule cells (38). Recent studies have demonstrated that mutations incomponents ofshh/ptcpathway can result in medulloblastoma forma-tion and other tumors. Loss ofptc expression, as a consequence eitherof germ-line mutation or somatic mutation in sporadic tumors, hasbeen linked to medulloblastoma. Furthermore, overexpression or ac-tivating mutations inshhalso lead to abnegation ofptc function andsubsequent tumorigenesis (39–41).
Mice heterozygous forptc, as a consequence of the targeted dis-ruption of one allele, develop brain tumors and provide a uniqueopportunity to investigate an animal model for medulloblastoma (24,25). Medulloblastomas that arise inptc1/2 mice express several of theintermediate filament proteins found previously to be present inhuman medulloblastoma cells. This is consistent with the proposedprimitive neuroectodermal cell of origin of the tumors. Tumor cellsretain the potential to differentiate along multiple pathways includingneuronal, glial, and ependymal cell lineages (38). Expression of NF,synaptophysin,trkC, and reelin as well as GFAP within the tumorsuggests that both neuronal and glial differentiation occurs. Here weshow that the normalptc allele is retained and expressed in medullo-blastomas arising inptc1/2 mice. We demonstrate at least two majorptc transcripts (approximately 8 and 12.5 kb) in all control and tumorsamples and 5 of 10 tumors also expressed a third, smaller transcriptof ;5 kb. At least fiveptc mRNA transcripts have been detected inhuman tissues (9), and the presence of the two largerptc transcripts incontrol as well as tumor tissues in the present study suggests thatseveralptc transcripts are also expressed by mouse tissues. The levelsof ptcexpression varied among the tumors, regardless of mouse strain,age at tumor detection, or anatomical location within the cerebellum.This demonstrates that complete loss ofptc expression is not aprerequisite for tumorigenesis. Thus,ptc is not acting as a classictumor suppressor gene in this mouse model.
Consistently,gli1, a candidate target gene thought to be activatedby thehh/ptcpathway, was found to be expressed at high levels in allof the mouse medulloblastomas analyzed (Figs. 3 and 4). This is anindication that there is increasedshhsignaling or reducedptc activityin the tumor cells (19). The expression levels ofgli1 were too low inthe normal tissues examined to determine whether loss of one allele ofptc caused up-regulation of expression. In contrast to the patterns ofgli1 expression, variable levels ofptc mRNA were observed amongthe tumors, and some showed only very low levels of expression.Previously, Ptc was reported to repress its own transcription, andreduced levels of Ptc activity were believed to increase transcriptionof ptc (42–44). This does not appear to be the case in mouse medul-loblastoma cells. In addition, there was approximately twice as muchptc mRNA in cerebellar extracts from normal mice compared withptc1/2 mice (Fig. 3), indicating that Ptc does not repress its ownexpression in mouse cerebellum. The regulatory pathway involvingptc is rather complex, and it is possible that there are additionaldefects in the tumor cells. Furthermore, the expression patterns of thegenes we examined may differ in the proposed granule cell precursorpopulation from which the tumors are believed to arise, comparedwith the postmitotic granule neurons present in the adult cerebellum.
Medulloblastomas inptc1/2 mice seem to arise from a focal lesion,and it is clear that many areas of the cerebellum have a relativelynormal histological appearance (Fig. 2D). Thus, ptc haploinsuffi-ciency alone is not sufficient to induce tumor formation, and it islikely that the tumors contain additional genetic lesions. Here wedemonstrate that the additional mutations do not involve the secondallele ofptc; however, it is still possible that other components of thepathway are mutated in the mouse tumors. The majority of human
Fig. 6. MaeII digest ofptc PCR products amplified from tail DNA from 129Sv,C57BL/6, andptc1/2 mice (nos. 199, 241, 448, and 646) that developed tumors usingprimers (ptcF8 corresponding to bases 3781–3804 andptcR8 complementary to bases4283–4305) that amplified a 524-bp fragment of ptc. Digestion of the 524-bp PCRproduct resulted in two fragments of 235 and 289 bp that migrate as a doublet on the 1%agarose gel. This presence of both 129 and C57BL/6 alleles was illustrated by comparisonof untreated (2) andMaeII-treated PCR products by agarose gel electrophoresis in thepresence of ethidium bromide. Mice 199 and 448 carried alleles from C57BL/6 and thetargeted 129Sv allele as indicated by the enzyme digest, whereas bothptcalleles in mouse241 were derived from 129Sv DNA.
medulloblastomas examined do not harbor mutations inptc, shh,orsmo(4–7, 45). It is possible that characterization of medulloblastomasfrom ptc1/2 mice may lead to the identification of genes that aremutated in the human disease.
crossed onto a CD-1 background. In addition to medulloblastomas,they observed an incidence soft tissue sarcomas, referred to as rhab-domyosarcoma, of 9%. In contrast, we detected a 4% incidence of softtissue sarcomas on the C57BL/6 background. The tumors appearedpale and fibrous on gross examination. The tumor cell types were notcharacterized in detail; therefore, we cannot say definitively whetherthey were rhabdomyosarcoma. The reasons for the different tumorspectra observed in the two studies are unclear. This may be areflection of the different strain background or because the strategyused to disruptptc by Hahnet al. (25) involved deletion of exons 6and 7, whereas the disruption reported by Goodrichet al. (24) re-moves most of exon 1 and all of exon 2.
Variable tumor incidences have been reported in different sub-strains of mice; for example, testicular teratoma is relatively rare inmost mouse strains, but an incidence of 1–3% has been reported in129Sv mice (46). Here we found no significant difference in theincidence of medulloblastomas on a 129Sv or C57BL/6 background.Surprisingly, we did encounter a high degree of DNA sequencepolymorphism ofptc between these two strains. In the 129Sv strain,one of the polymorphisms results in a methionine at amino acidresidue 1339 compared with a threonine in C57BL/6 mice. Thisresidue is located in the terminal 100 amino acids of the Ptc protein ina region of Ptc that interacts with Smo (47, 48).
Complete inactivation ofptc appears to be an important step in thedevelopment of adult basal cell carcinoma and in a subset of humanmedulloblastomas (5–8). However, our data demonstrate that braintumors of similar phenotype and anatomical location arise in micewith no loss of heterozygosity of theptc locus. Similarly, in medul-loblastomas from individuals with BCNS, the secondptc allele wasreported to be mutated in only one of three cases examined (12). Inapproximately one-half of the sporadic medulloblastomas analyzed inwhich 9q loss or mutation ofptcwas demonstrated, no mutations werefound in the secondptc allele (4–8). Persistence of the wild-typeptcallele and variable levels ofptc mRNA expression were detected inmany tumors (4, 7). This raises the question: what is the functionalconsequence ofptc haploinsufficiency that leads to medulloblastomaformation? One possibility is that the reduced level of Ptc expression,caused by loss of one allele, results in increased proliferation ofgranule cell precursors. The increase in cell proliferation may bebalanced by elevated levels of cell death such that the total number ofgranule neurons does not change. In fact, during normal brain devel-opment there is an overproduction of neurons and a subsequent loss ofexcess cells by apoptosis. The increased turnover of granule cellprecursors would enhance the possibility that additional mutationswould arise in these cells. These mutations could either promotecontinued cell division or escape from cell death, resulting in tumor-igenesis. Hopefully, continued analysis ofptc1/2 mice will elucidatethe molecular and cellular mechanisms underlying formation ofmedulloblastoma in mice and humans.
ACKNOWLEDGMENTS
We thank M. Scott and L. Goodrich for theptc1/2 mice andptc plasmids(617 and M2-3); A. Ruiz i Altaba for the humangli1 plasmid; A. Joyner andF. Lamballe for thegli3 and trkC plasmids; B. Bain for attentive care of themice; and R. L. Johnson for helpful discussions.
REFERENCES
1. Russel, D. S., and Rubinstein, L. J. Pathology of Tumors of the Nervous System, Ed.6, Vol. I. New York: Oxford University Press, 1998.
2. Reardon, D. A., Michalkiewicz, E., Boyett, J. M., Sublett, J. E., Entrekin, R. E.,Ragsdale, S. T., Valentine, M. B., Behm, F. G., Li, H., Heideman, R. L., Kun, L. E.,Shapiro, D. N., and Look, A. T. Extensive genomic abnormalities in childhoodmedulloblastoma by comparative genomic hybridization. Cancer Res.,57: 4042–4047, 1997.
3. Biegel, J. A. Genetics of pediatric central nervous system tumors. J. Pediatr. Hema-tol./Oncol.,19: 492–501, 1997.
4. Pietsch, T., Waha, A., Koch, A., Kraus, J., Albrecht, S., Tonn, J., Sorensen, N.,Berthold, F., Henk, B., Schmandt, N., Wolf, H. K., von Deimling, A., Wainwright, B.,Chenevix-Trench, G., Wiestler, O. D., and Wicking, C. Medulloblastomas of thedesmoplastic variant carry mutations of the human homologue ofDrosophilapatched. Cancer Res.,57: 2085–2088, 1997.
5. Raffel, C., Jenkins, R. B., Frederick, L., Hebrink, D., Alderete, B., Fults, D. W., andJames, C. D. Sporadic medulloblastomas contain PTCH mutations. Cancer Res.,57:842–845, 1997.
6. Vorechovsky, I., Tingby, O., Hartman, M., Stromberg, B., Nister, M., Collins, V. P.,and Toftgard, R. Somatic mutations in the human homologue ofDrosophila patchedin primitive neuroectodermal tumours. Oncogene,15: 361–366, 1997.
7. Wolter, M., Reifenberger, J., Sommer, C., Ruzicka, T., and Reifenberger, G. Muta-tions in the human homologue of the Drosophila segment polarity genepatched(PTCH) in sporadic basal cell carcinomas of the skin and primitive neuroectodermaltumors of the central nervous system. Cancer Res.,57: 2581–2585, 1997.
8. Xie, J., Johnson, R. L., Zhang, X., Bare, J. W., Waldman, F. M., Cogen, P. H., Menon,A. G., Warren, R. S., Chen, L. C., Scott, M. P., and Epstein, E. H., Jr. Mutations ofthePATCHEDgene in several types of sporadic extracutaneous tumors. Cancer Res.,57: 2369–2372, 1997.
9. Hahn, H., Wicking, C., Zaphiropoulous, P. G., Gailani, M. R., Shanley, S.,Chidambaram, A., Vorechovsky, I., Holmberg, E., Unden, A. B., Gillies, S., Negus,K., Smyth, I., Pressman, C., Leffell, D. J., Gerrard, B., Goldstein, A. M., Dean, M.,Toftgard, R., Chenevix-Trench, G., Wainwright, B., and Bale, A. E. Mutations of thehuman homolog ofDrosophila patchedin the nevoid basal cell carcinoma syndrome.Cell, 85: 841–851, 1996.
10. Johnson, R. L., Rothman, A. L., Xie, J., Goodrich, L. V., Bare, J. W., Bonifas, J. M.,Quinn, A. G., Myers, R. M., Cox, D. R., Epstein, E. H., Jr., and Scott, M. P. Humanhomolog ofpatched,a candidate gene for the basal cell nevus syndrome. Science(Washington DC),272: 1668–1671, 1996.
11. Cogen, P. H., and McDonald, J. D. Tumor suppressor genes and medulloblastoma.J. Neuro-Oncol.,29: 103–112, 1996.
12. Cowan, R., Hoban, P., Kelsey, A., Birch, J. M., Gattamaneni, R., and Evans, D. G.The gene for the naevoid basal cell carcinoma syndrome acts as a tumour-suppressorgene in medulloblastoma. Br. J. Cancer,76: 141–145, 1997.
13. Nusslein-Volhard, C., and Wieschaus, E. Mutations affecting segment number andpolarity in Drosophila. Nature (Lond.),287: 795–801, 1980.
14. Hooper, J. E., and Scott, M. P. TheDrosophila patchedgene encodes a putativemembrane protein required for segmental patterning. Cell,59: 751–765, 1989.
15. Murone, M., Rosenthal, A., and de Sauvage, F. J. Sonic hedgehog signaling by thepatched-smoothened receptor complex. Curr. Biol.,9: 76–84, 1999.
16. Ingham, P. W. Transducing Hedgehog: the story so far. EMBO J.,17: 3505–3511,1998.
17. Tabin, C. J., and McMahon, A. P. Recent advances in hedgehog signalling. TrendsCell Biol., 7: 442–446, 1997.
18. Roessler, E., Ward, D. E., Gaudenz, K., Belloni, E., Scherer, S. W., Donnai, D.,Siegel-Bartelt, J., Tsui, L. C., and Muenke, M. Cytogenetic rearrangements involvingthe loss of the Sonic Hedgehog gene at 7q36 cause holoprosencephaly. Hum. Genet.,100: 172–181, 1997.
19. Goodrich, L. V., and Scott, M. P.Hedgehogandpatchedin neural development anddisease. Neuron,21: 1243–1257, 1998.
20. Kinzler, K. W., Bigner, S. H., Bigner, D. D., Trent, J. M., Law, M. L., O’Brien, S. J.,Wong, A. J., and Vogelstein, B. Identification of an amplified, highly expressed genein a human glioma. Science (Washington DC),236: 70–73, 1987.
21. Kinzler, K. W., Ruppert, J. M., Bigner, S. H., and Vogelstein, B. TheGLI gene is amember of the Kruppel family of zinc finger proteins. Nature (Lond.),332: 371–374,1988.
22. Gorlin, R. J. Nevoid basal-cell carcinoma syndrome. Medicine,66: 98–113, 1987.23. Kimonis, V. E., Goldstein, A. M., Pastakia, B., Yang, M. L., Kase, R., DiGiovanna,
J. J., Bale, A. E., and Bale, S. J. Clinical manifestations in 105 persons with nevoidbasal cell carcinoma syndrome. Am. J. Med. Genet.,69: 299–308, 1997.
24. Goodrich, L. V., Milenkovic, L., Higgins, K. M., and Scott, M. P. Altered neural cellfates and medulloblastoma in mouse patched mutants. Science (Washington DC),277: 1109–1113, 1997.
25. Hahn, H., Wojnowski, L., Zimmer, A. M., Hall, J., Miller, G., and Zimmer, A.Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlinsyndrome [see comments]. Nat. Med.,4: 619–622, 1998.
26. Aszterbaum, M., Rothman, A., Johnson, R. L., Fisher, M., Xie, J., Bonifas, J. M.,Zhang, X., Scott, M. P., and Epstein, E. H., Jr. Identification of mutations in thehumanPATCHEDgene in sporadic basal cell carcinomas and in patients with thebasal cell nevus syndrome. J. Investig. Dermatol.,110: 885–888, 1998.
27. Hui, C. C., Slusarski, D., Platt, K. A., Holmgren, R., and Joyner, A. L. Expression ofthree mouse homologs of the Drosophila segment polarity genecubitus interruptus,Gli, Gli-2, andGli-3, in ectoderm- and mesoderm-derived tissues suggests multipleroles during postimplantation development. Dev. Biol.,162: 402–413, 1994.
28. D’Arcangelo, G., Miao, G. G., Chen, S. C., Soares, H. D., Morgan, J. I., and Curran,T. A protein related to extracellular matrix proteins deleted in the mouse mutant reeler[see comments]. Nature (Lond.),374: 719–723, 1995.
29. Lamballe, F., Klein, R., and Barbacid, M. trkC, a new member of the trk family oftyrosine protein kinases, is a receptor for neurotrophin-3. Cell,66: 967–979, 1991.
30. Dahmane, N., Lee, J., Robins, P., Heller, P., and Ruiz i Altaba, A. Activation of thetranscription factor Gli1 and the Sonic hedgehog signalling pathway in skin tumours[published erratum appears in Nature (Lond.),390: 536, 1997]. Nature (Lond.).,389:876–881, 1997.
31. Roberts, W. M., Douglass, E. C., Peiper, S. C., Houghton, P. J., and Look, A. T.Amplification of thegli gene in childhood sarcomas. Cancer Res.,49: 5407–5413, 1989.
32. Xanthoudakis, S., Miao, G., Wang, F., Pan, Y. C., and Curran, T. Redox activation ofFos-Jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J.,11:3323–3335, 1992.
33. Segal, R. A., Goumnerova, L. C., Kwon, Y. K., Stiles, C. D., and Pomeroy, S. L.Expression of the neurotrophin receptor TrkC is linked to a favorable outcome inmedulloblastoma. Proc. Natl. Acad. Sci. USA,91: 12867–12871, 1994.
34. Jacobson, M. Developmental Neurobiology, Ed. 3. New York: Plenum PublishingCorp., 1991.
35. Wechsler-Reya, R. J., and Scott, M. P. Control of neuronal precursor proliferation inthe cerebellum bySonic Hedgehog[see comments]. Neuron,22: 103–114, 1999.
36. Rowitch, D. H., St-Jacques, B., Lee, S. M. K., Flax, J. D., Snyder, E. Y., andMcMahon, A. P.Sonic hedgehogregulates proliferation and inhibits differentiation ofCNS precursor cells. J. Neurosci.,19: 8954–8965, 1999.
37. Hatten, M. E., and Heintz, N. Mechanisms of neural patterning and specification inthe developing cerebellum. Annu. Rev. Neurosci.,18: 385–408, 1995.
38. Rorke, L. B., Trojanowski, J. Q., Lee, V. M., Zimmerman, R. A., Sutton, L. N.,Biegel, J. A., Goldwein, J. W., and Packer, R. J. Primitive neuroectodermal tumors ofthe central nervous system. Brain Pathol.,7: 765–784, 1997.
39. Fan, H., Oro, A. E., Scott, M. P., and Khavari, P. A. Induction of basal cell carcinomafeatures in transgenic human skin expressingSonic Hedgehog. Nat. Med.,3: 788–792, 1997.
40. Dahmane, N., and Ruiz-i-Altaba, A.Sonic hedgehogregulates the growth andpatterning of the cerebellum. Development (Camb.),126: 3089–3100, 1999.
41. Oro, A. E., Higgins, K. M., Hu, Z., Bonifas, J. M., Epstein, E. H., Jr., and Scott, M. P.Basal cell carcinomas in mice overexpressingsonic hedgehog. Science (WashingtonDC), 276: 817–821, 1997.
42. Capdevila, J., Estrada, M. P., Sanchez-Herrero, E., and Guerrero, I. The Drosophilasegment polarity genepatchedinteracts with decapentaplegic in wing development.EMBO J.,13: 71–82, 1994.
43. Marigo, V., and Tabin, C. J. Regulation ofpatchedby sonic hedgehogin thedeveloping neural tube. Proc. Natl. Acad. Sci. USA,93: 9346–9351, 1996.
44. Goodrich, L. V., Johnson, R. L., Milenkovic, L., McMahon, J. A., and Scott, M. P.Conservation of the hedgehog/patched signaling pathway from flies to mice: induc-tion of a mousepatchedgene byHedgehog. Genes Dev.,10: 301–312, 1996.
45. Zurawel, R. H., Allen, C., Chiappa, S., Cato, W., Biegel, J., Cogen, P., de Sauvage,F., and Raffel, C. Analysis of PTCH/SMO/SHH pathway genes in medulloblastoma.Genes Chromosomes Cancer,27: 44–51, 2000.
46. Simpson, E. M., Linder, C. C., Sargent, E. E., Davisson, M. T., Mobraaten, L. E., andSharp, J. J. Genetic variation among 129 substrains and its importance for targetedmutagenesis in mice. Nat. Genet.,16: 19–27, 1997.
47. Chen, Y., and Struhl, G. Dual roles forpatchedin sequestering and transducingHedgehog. Cell,87: 553–563, 1996.
48. Stone, D. M., Hynes, M., Armanini, M., Swanson, T. A., Gu, Q., Johnson, R. L.,Scott, M. P., Pennica, D., Goddard, A., Phillips, H., Noll, M., Hooper, J. E., deSauvage, F., and Rosenthal, A. The tumour-suppressor genepatchedencodes acandidate receptor forSonic hedgehog. Nature (Lond.),384: 129–134, 1996.
![Page 1: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/1.jpg)
![Page 2: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/2.jpg)
![Page 3: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/3.jpg)
![Page 4: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/4.jpg)
![Page 5: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/5.jpg)
![Page 6: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/6.jpg)
![Page 7: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/7.jpg)
![Page 8: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/8.jpg)
![Page 9: The Normal patched Allele Is Expressed in Medulloblastomas ...cancerres.aacrjournals.org/content/60/8/2239.full.pdf · [CANCER RESEARCH 60, 2239–2246, April 15, 2000] The Normal](https://reader034.documents.pub/reader034/viewer/2022042707/5a74e6ed7f8b9a93088bf69f/html5/thumbnails/9.jpg)
