[CANCERRESEARCH54, 649-652, February 1, 19941

Advances in Brief

Analysis of the p53 Gene and Its Expression in Human Glioblastoma Cells1

Erwin G. Van Meir,2 Tetsuro Kikuchi, Mitsuhiro Tada, Hong Li, Annie-Claire Diserens, Brian E. Wojcik,

H-J. Su Huang, Theodore Friedmann, Nicolas de Thbolet, and Webster K. CaveneeLaboratory of Tumor Biology and Genetics, Neurosurgery Service, University Hospital (CHUV), rue de Bugnon 5, lOll Lausanne, Switzerland fE. G. V M., T. K., M. T., H. L.,A-C. D., N. d. TJ; Ludwig Institute for Cancer Research fE. G. V M., H-J. S. H., W. K. C.J Departments of Pediatrics fB. E. W, T F/ and Medicine fH-J. S. H., W K. C./ andthe Center for Molecular Genetics fT F, W@K. C.J, University of California, San Diego, California 92093

Abstract

Chromosome l7p has been shown to be an early and frequent target forloss of heterozygosity through mitotic recombination in astrocytomas.These losses are frequently accompanied by point mutations in the p13gene of the remaining allele, resulting in loss of wild type p5.3 function.However, a fraction of astrocytomas retain constitutional heterozygosityand do not have p5.3 mutations; some of these lose wild type p53 activitythrough binding to the protein product of amplified indm2 genes. To test

whether loss of wild type p53 biological function is a necessary step inastrocytoma progression we analyzed p53 expression and biological function in 13 glioma cell lines. All the cell lines expressed a 2.8-kilobase p5.3transcript and showed various amounts of p53 protein by immunoprecipitation, except for cell line LN-Z308 which had only a small truncatedp53 mRNA and no protein expression. To test whether the p53 expressedin these cell lines was functionally wild type or mutant we transfectedthem with a plasmid construct harboring a chloramphenicol acetyltransferase (CAT) reporter gene under the control of transcriptional elementsthat are induced by wild type but not mutant p53. Four lines were shownto retain wild typep53 function. Sequencing ofthep53 gene in two of thesecell lines confirmed the wild type genotype. These results show that mactivation of the p5.3 gene is not an obligatory step in glioblastoma genesis.This suggests either that two pathways (p53 inactivation dependent orindependent) may lead to a tumor group classified histologically as glioblastoma or that in some cases p5.3 mutations are bypassed due to thepresence of mutations in downstream effector genes.

Introduction

The most frequent tumors of central nervous system neuroepithelialorigin are astrocytomas and glioblastomas which comprise about 25—30% and 50% of cerebral gliomas, respectively. The presence ofanaplastic foci in benign astrocytomas and the propensity of thesetumors to recur after surgery with increased malignancy lead to theconcept that they undergo progressive malignant transformation. Thisis believed to start with a series of initiation events in an astrocyte orastrocyte precursor cell, leading to a benign astrocytoma, followed bythe progressive accumulation of new genetic defects causing anaplasiaand, finally, a glioblastoma (1). However, it is not clear whether alltumors classified histologically as glioblastoma fit this progressivemodel of astrocytic evolution toward malignancy. It is well documented that a category of glioblastomas appears de novo with afulminant clinical course. These tumors, while indistinguishable at themorphological level, do not show histological evidence for evolutionfrom one of the more differentiated forms of glioma and emphasizethe need for the development of better criteria for subclassification

(1).

Received 11/29/93; accepted 12/17/93.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported in part by Advanced Fellowship 823A-030692 (E. G. V. M.)

and Grant 3.595.087 (N. d. T.) of the Swiss National Science Foundation and by a “PaoloBaffi―Fellowship (E. G. V.M.) from the “Formazioneper Ia Formazione Oncologica―andthe European Institute of Oncology.

2 To whom requests for reprints should be addressed at, Laboratory of Tumor Biology

and Genetics, Service of Neurosurgery, University Hospital, (CHUV) rue du Bugnon 5,CH-10l 1 Tausanne, Switzerland.

The application of molecular genetics has shown that human tumordevelopment can occur with an accumulation of stage-specific geneticalterations (2, 3). This model implies that the more malignant stageswould have retained the genetic defects which occurred in the earlierstages in addition to acquiring new genetic defects. In support of sucha model for astrocytoma, alterations of the p53 gene were found in

about 30% of benign and anaplastic astrocytomas as well as glioblastomas (reviewed in Refs. 4 and 5). Further evidence came from thestatus of the p53 gene in patients with recurrent tumors (6, 7) where

it was shown that when no p53 mutation was detected in the primary

tumor, a de novo p53 mutated allele appeared in the recurrent tumor

and that these new mutations were already present in anaplastic fociwithin the indolent benign tumor. Additionally, in a patient having a

p53 mutation of one allele in his primary tumor, it was shown that theremaining wild type allele had acquired an independent p53 mutationafter recurrence (7). If the pathway of malignant transformation inastrocytoma obligatorily involves the elimination of p53 activity, allsuch tumors would be expected to contain mutated or inactiveWTp53.3

To test this hypothesis, we analyzed p53 expression at the structuraland functional level in a large series of glioblastoma cell lines. Wereasoned that such cell lines would likely be representative of the

variable ontogenics of glioblastoma and that their use would allow us

to perform functional studies that are not possible with primary tumorsamples. We also compared the p.5.3status ofthe cells with their abilityto grow as s.c. tumors in immunodeficient mice. Our results show that4 of 13gliomacell linesexpressedwildtypep53 as measuredbytranscriptional activation of a WTp53-dependent promoter. We foundno correlation between the p53 status of these cell lines and theirability to form s.c. tumors in nu/nu mice.

Materials and Methods

Cell Lines. The glioma cell lines used in this study and their cultureconditions have been described previously (Ref. 8 and American Type Culture

Collection). All cell lines were derived from glioblastomas, except two fromastrocytomas (LN-319 and U373MG) and one from a gliosarcoma (D247MG).

RNA Isolation and Hybridization. Total RNA was isolated from cell linesor solid tumors and used for Northern blotting experiments with randomlabeled probes as described previously (9). The 1584-base pair Aval/XbaIfragment of the p5.3 coding region of plasmid php53B was used as a probe forPS3, and the PstI 1100-base pair fragment of plasmid pAL41 was used as a

probe for @3-actin(9).Amplification and Sequencing of the p53 Locus. One @gof genomic

DNA of cell lines LN-18, LN-229, LN-319, and LN-428 was amplified byPCR using primers p53IN4ALand p53IN7BL for exons 5, 6, and 7 and primersp53IN7AL and p53IN9BL for exons 8 and 9 as described previously (10). Foreach cell line the DNA of three independent PCR reactions was mixed, theamplified fragment was cloned in pBluescript (Stratagene), and 4 to 10 mdividual clones were sequenced as described (10). For cell line U87MG, se

3 The abbreviations used are: WTpS3, wild type p53 protein; CAT, chloramphenicol

acetyltransferase; PCR, polymerase chain reaction; PBS, phosphate-buffered saline; SDS,sodium dodecyl sulfate; i.e., intracranial.

649

Research. on December 9, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

p53 GENE EXPRESSIONIN GUOBLASTOMACELL LINES

quencing of exons 2 to 11 of the p5.3 gene was performed by BioServeBiotechnologies, Inc. (Laurel, MD) using direct PCR sequencing on genomicDNA.

‘fransientTransfection Bioassay. PGI3CATand MG15CATplasmids (11)were introduced into recipient cell lines either by electroporation or by calciumphosphate transfection as described (12). Forty-eight h after transfection, cellswere scraped off the dishes and cell lysates were tested for the presence of CATactivity. CAT reaction products were separated from the [‘4C]chloramphenicolsubstrate using xylene extraction and were counted in a liquid scintillationcounter as described (13). Each of these experiments was replicated at leasttwice and each point was in triplicate. As a control for false negative results,we performed separate cotransfections with plasmids pc53-SN3 (expressingWTp53) and PG,3CAT in these cell lines and demonstrated activation of the

CAT reporter gene for all cell lines. The background activity obtained in thesame cells by cotransfection of the pc53-SN3 plasmid and the MG15CATcontrol plasmid was subtracted from the cpm obtained by transfection ofplasmid PG13CATalone or cotransfected with pc53-SN3 in each cell line.Then, the relative trans-activation activity of PG13CATin each line wascalculated as a percentage of the maximal activation obtained when PG13CAT

was cotransfected with pc53-SN3. Transactivation was considered positive (+)when the relative trans-activation activity was above 10%.

Determination of Thmorigenicity in Nude Mice. Female nu/nu miceweighing 20—22g were housed in sterile conditions with free access to foodand water. Room temperature (22°C),humidity (55—65%)and a 12-h light,12-h dark cycle were kept constant. Cells for transplantation were harvested bytrypsinization, washed in PBS, and resuspended in Dulbecco's modifiedEagle's medium without serum. Cells (10@-5X 10') were injected at a singles.c. site into each mouse at 2—3 months of age. All mice were examined

regularly for development of tumors. Cells were scored as tumorigenic if apalpable nodule appeared at the site of injection within 3 months.

Immunoprecipitation. For metabolic labeling, 50—90%confluent cellswere washed with PBS, grown for 1 h in methionine-free minimal essentialmedium with 10% dialyzed serum (Sigma F0392; Sigma, St. Louis, MO), and

then further incubated for 90 mm in the same medium with 150 @Ci/mlE35Slmethionine (Amersham, Arlington, IL). Subsequently, the cells werewashed with PBS and lysed for 10 mm at 4°Cin PBS containing 1% Tritonx-100, 0.5% sodium deoxycholate,0.1% SDS, 0.2% sodium azide, and0.004% sodium fluoride supplemented with the following protease inhibitors:5 @Wmlantipain,50 @iWmlaprotinin,5 p@g/mlleupeptin, 5 @g/mlpepstatinA,and 2 msi phenylmethylsulfonyl fluoride (Sigma). The cells were scraped offthe plates with a rubber policeman and subjected to 3 freeze-thaw cycles, andthe lysate was filtered through a 0.22 @mMillex filter (Millipore). An aliquotof each lysate was taken and the activity was measured in a Beckmannscintillation counter. One @gof anti-p53 monoclonal antibody G59-12 (Pharmingen, San Diego, CA) and 70 pAof a 50% solution of protein A-SepharoseCL-4B (Pharmacia, NJ) in H2O saturated with 4% bovine serum albumin wereadded to equal activities of labeled lysates, and the mixture was left forimmunoprecipitation overnight at 4°Con a rotating wheel. The Sepharosebeads were then washed 7 times in PBS containing 1% Triton X-100, 0.5%sodium deoxycholate, 0.1% SDS, 0.2% sodium azide, and 0.004% sodiumfluoride. The immunoprecipitated proteins were recovered from the beads byheating for 3 mm at 95°Cin loading buffer and electrophoresed through a 10%SDS-polyacrylamide gel. The gel was dried and exposed to X-ray film.

Results and Discussion

The expression of the p53 gene was determined by Northern blotting experiments of cytoplasmic RNA extracted from 16 glioma celllines. All cell lines expressed the 2.8 kilobase p53 mRNA albeit atvariable levels; the exception was LN-Z308 which expressed a smallamount of a truncated p53 transcript (Fig. 1). Various levels of p53protein were detected by immunoprecipitation in 18 of 19 cell linestested and as expected no p53 protein could be detected in the LNZ308 cell line; substantiating the mRNA data (Fig. 2).

In order to evaluate whether the proteins expressed by a subset ofthese cell lines had wild type or mutant transactivation activities, weperformed transient DNA transfection experiments with the reporterplasmid PG13-CAT (11). The polyomavirus early promoter in this

construct is flanked by 13 repeats of a p53 binding sequence (PG) thatallows WT, but not mutant, p53 transcriptional activation of a CAT

288- .@ .@ai@

@e.... .e.*•@.. -2.8Kb188- . . S

. I

650

(!@ (.@@ csJ c'),-c'),-.@

@c?c?c')@@ .z,@zzzc@zzz±@zzzz@z

@_J_J_J_J_J_J@

@a. @@jfl

Fig. 1. p53 mRNA detection in glioblastoma cell lines. Northern blot analysis wasperformed on 10 lABof cytoplasmic RNA from each cell line using a p53 complementaryDNA probe or a @3-actincomplementary DNA probe as a control for loading amounts andRNA integrity. RNA from a primary culture of human scalp fibroblasts (445Fbl.) was alsoused. Kb, kilobases.

!@@@@@@@ r@@@

p53@@

Fig. 2. p53 protein detection in glioblastoma cell lines. Immunoprecipitation analysiswas performed with anti-p53 monoclonal antibody G59-12 on equal activities of labeledlysates obtained from metabolically labeled proteins of each cell line as described in“Materialsand Methods.―The slow and fast migrating forms of the p53 protein areindicated. Overexposure of this blot allowed the detection of p53 protein in Hs683 cellsbut not in LN-Z308 cells.

reporter gene and has been validated previously in the colon carcinoma cell line HCfll6 for the ‘43Val-Ala,‘75Arg-His,248Arg-Trpand 273Arg-His mutants (11). The majority of glioma cell lines weanalyzed (9 of 13) express p53 protein which is nonfunctional in thisassay (Table 1). This result is consistent with the expectation of a lackof activity for cell lines T98G, U251MG and U373MG which werepreviously shown to contain mutant p53 alleles (14, 15). In order todetermine whether the lack of transcriptional activation in the othercell lines was due to the presence in them of p53 alleles, we determined their nucleotide sequence (exons 5—9)in three additional celllines and showed that they also contained mutated p53 alleles(Table 1).

A minority (4 of 13) of the glioma cell lines (U87MG, D247MG,

LN-229, and U138MG) had demonstrable trans-activation activityusing the PGI3CAT reporter, but not with MG15-CAT (a constructsimilar to PG13CAT, but with a mutated oligonucleotide repeat whichdoes not bind WTp53 in vitro), showing that they contain functionalWTp53 activity. To corroborate this interpretation, the entire codingregion (exons 2—li) of the p53 gene in cell line U87MG was sequenced and determined to be wild type. This latter observationcomplements previous studies (16) which had shown that exons 4—9

of the p53 gene in D247MG were wild type in sequence. Sequencingof exons 5—8of the p53 gene in cell line LN-229 showed the presenceof two alleles; one was wild type (6 of 8 clones sequenced) and thesecond carried an A to G transition at amino acid 164 (2 of 8 clonessequenced). This leads to a lysine to glutamine amino acid changeoutside the evolutionarily conserved regions of the p53 gene. We arefurther examining whether this change is an artefact, a polymorphismin the p53 gene, or whether it represents a recessive p.5.3 mutation.Since these cell lines can genetically drift over time, we cannot ex

Research. on December 9, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

Cell linesPG13-CAT transactivationp53 genotypep53

sequencechangeCodonMutation

(or polymorphism)Amino acidTumorigenicityin nu/numiceLN-18

LN-229

LN-Z308LN-319LN-428

U87MGU138MGU2SIMGU373MGD247MGT98GHs683A172

a These mutations—

(0)

+ (77)

— (0)

— (0)

— (2.2)

+ (—100)+ (12.7)

- (0)

- (0.3)

+ (—100)— (0.1)

- (0)

— (0.7)

may be in the same allele; tHeterozygous

Heterozygous

HomozygousHomozygousHeterozygous

HomozygousNDNANAHomozygousHomozygousNDHeterozygous

he sequencing cannot determinate.238

164

/175173―282―

//

273273/

237//TGT

-s TCF@VT(exons 5—9)

AAG —‘GAGWT (exons 5—8)RearrangementCGC —aCACGTG—‘ATGCGG—TGGWT (exons 2—11)NDCGT -+ CATCGT -@CAT‘NT(exons 4—9)ATG—+ATANDNDCys

—‘5cr

Lys —‘Glu

/Mg —aHisVal—@MetArg-@Trp

//

Arg -â€HisArg —@His

/Met --â€Ile

//T

T

TTNT

TNTTTNTNTNTNT

p53 GENE EXPRESSIONIN GLIOBLASTOMACELL LINES

Table 1 p53 status in glioma cell lines and their ability to form s.c. tumors in nude mice.The positive (+) or negative (—)results of the transient transfection assays with plasmid PG13CATare indicated for each cell line. Transcriptional activation of the WTp53-dependent

promoter of this plasmid was measured by chloramphenicol acetyltransferase activity and is indicated in parentheses as a percentage of the maximal activation obtained in the cellswhen this plasmid is cotransfected with pc53-SN3, a plasmid expressing wild type p53 (see “Materialsand Methods―).The p53 genotype was considered heterozygous if two typesof alleles were detected by sequencing or by migration on SDS-polyacrylamide gel electrophoresis gels (LN-428 and A172; see Fig. 1) and homozygous when there was evidence foronly one type of allele. The p53 codons involved and the changes in nucleotide (boldface print) and amino acid composition are indicated; 6 of 9 clones sequenced were mutant atcodon 238 in cell line LN-18, 2 of 8 clones at codon 164 in cell line LN-229, 10 of 10 clones at codon 175 in cell line LN-319, and 7 of 10 clones at codon 173 and 3 of 4 clonesat codon 281 in cell line LN-428. Mutations in the p53 gene had previously been found by sequencing for cell lines T98G (14, 28), U2S1MG, and U373MG (28). No p53 mutationswere found in cell lines D247MG (16) and U87MG. The ability ofthese cell lines to form s.c. tumors in nu/nu mice and also by stereotactic injection into the nu/nu mouse brain striatumin some cases (LN-Z308, U87MG, T98G) is indicated by a T (tumor) or NT (no tumor). The maximum number of cells injected was 5 X 10@(s.c.) and 5 X lO@(in the brain). NA,not available; ND, not done; /, not relevant.

dude the possibility that a small proportion of cells might containalleles different than those assessed here.

These results, in combination with other analyses of p5.3 in astrocytoma (reviewed in Ref. 4), lend support to the hypothesis thatdifferent mechanisms may account for WTp53 inactivation in glioblastoma. One mechanism would involve the loss of one allele, possibly through mitotic recombination, and inactivation ofthe second bya point mutation, as represented by the LN-319 and T98G lines.Another possibility is the occurrence of a point mutation in one allelefollowed by an independent point mutation, as represented by cell lineLN-428. This further supports the idea that loss of WTp53 suppressorgene activity may occur progressively (7), the first step being themutation or loss of one allele which confers a small advantage to thesecells which is amplified upon the second alteration. Alternatively, celllines having both WFpI3 and mutated p53 alleles might be indicativeof single step WTp53 inactivation, the mutation in one allele conferring a dominant negative function, and such may be the situation withthe LN-18 cell line. Such a mechanism could be mimicked by theinactivation of wild type p53 protein through, for example, binding tothe mdm2 protein (5). Finally, the results obtained with cell lineLN-Z308 suggest that a fourth mechanism may also occur: loss of onep53 allele by rearrangementofthep53 gene and subsequentloss of thesecond allele, most likely by mitotic recombination.

In contrast, glioma cell lines U87MG, D247MG, LN-229, andU138MG retained the ability to express p53 proteins with wild typetransactivation activities. Two of these cell lines were previouslythought to be p53 deficient as shown by the lack of p53 mRNAexpression in U87MG (15) or p53 protein expression by Western blotin D247MG (16). This discrepancy with the present data may beexplained by the more sensitive immunoprecipitation and bioassay

used here that permit the detection of the short lived wild type p53protein. Sequencing of the hot spot regions ofp53 alleles in glioblastomas had already shown the absence of mutations in certain tumors(reviewed in Ref. 4), but this structural evidence could not conclusively demonstrate that the p53 proteins were active as other mechanisms may have prevented their normal function (17). Indeed, it hasbeen shown that in cancer cells functional p53 can be inactivated byother gene products such as mdm2 or E6 (reviewed in Ref. 18).

651

Recently, a small fraction (—8—10%)of glioblastomas without p53mutations were shown to contain amplification of the mdm2 gene (5).

In order to determine whether the p53 allelic status was correlatedwith the ability of the cells to form s.c. tumors in immunodeficientmice, 5 X 106 and 5 X i0@ cells were injected into nu/nu mice andevaluated for tumor formation after 3 months (Table 1). The presentresults indicate that the occurrence of p53 mutations is not necessary(U87MG expresses WTp53 and is tumorigenic) or sufficient (T98Gexpresses mutantp53 and is nontumorigenic) to induce tumorigenicityin nude mice. Whether this is representative of the in vivo situation inhumans or reflects an inadequacy in the assay is not clear. However,it is notable that limited comparisons of s.c. and i.c. implantations (wehave tested LN-Z308, T98G, and U87MG) gave qualitatively similarresults.

The presence of functional WTpS3 in some glioblastoma cell linesmay have implications with regard to the genetic subclassification ofglioblastomas which are a heterogeneous group of tumors (1, 19). Forexample, it may be that two oncogenic pathways can lead to a tumorclassified as glioblastoma: one is characterized by the inactivation of

WTpS3 function through allele loss, mutations, or binding to otherproteins such as mdm2; while the other is independent of the mactivation of the p13 gene function. It has been demonstrated that p53mutations can be a feature of the subgroup of glioblastomas that arethe end stage of the progression towards malignancy of astrocytomas(7). It remains to be tested whether glioblastomas arising in the absence of p53 inactivation also arise through a similar progressionscheme. In addition, it will be of interest to see whether this geneticdifference correlates with a specific clinical behavior such as those denovo glioblastomas with a fulminant clinical course (1) or patientswith long postoperative survival (10). In this context it is interestingto mention that a recent study on a limited number of patients suggeststhat glioblastomas with p53 mutations may be a subset that are morecommon in women and in younger patients (4).

The absence of mutated p53 alleles has been shown by sequencingin other tumor types, suggesting the existence of separate prognosticpatient groups (20—24),but p53 function was not tested in thesestudies. Despite the generality of p53 involvement in cancer, supportfor the occurrence ofp53 inactivation-independent oncogenic mecha

Research. on December 9, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

p53 GENE EXPRESSIONIN GUOBLASTOMA CELL UNES

Rca.,53: 957—961,1993.11. Kern, S. E., Pietenpol, J. A., Thiagalingam, S., Seymour, A., Kinzler, K. W., and

Vogelstein, B. Oncogenic forms ofp53 inhibitp53-regulated gene expression. Science(Washington DC), 256: 827—830,1992.

12. Graham, F. L., and Van der Eb, A. J. A new technique for the assay of infectivity ofhuman adenovirus 5 DNA. Virology, 52: 456-467, 1973.

13. Seed, B., and Sheen, J. Y. A simple phase-extraction assay for chioramphenicolacetyltransferase activity. Gene, 67: 271—277,1988.

14. Ullrich, S. J., Mercer, W. E., and Appella, E. Human wild-type p53 adopts a uniqueconformational and phosphorylation state in vivoduring growth arrest of glioblastomacells. Oncogene, 7: 1635—1643,1992.

15. Godbout, R., Miyakoshi, J., Dobler, K. D., Andison, R., Matsuo, K., Allalunis, T. M.,Takebe, H., and Day, R., III. Lack of expression of tumor-suppressor genes in humanmalignant glioma cell lines. Oncogene, 7: 1879—1884,1992.

16. Nigro, J. M., Baker, S. J., Preisinger, A. C., Jessup, J. M., Hostetter, R., Cleary, K.,Bigner, S. H., Davidson, N., Baylin, S., Devilee, P., et aL Mutations in the p53 geneoccur in diverse human tumour types. Nature (Lond.), 342: 705—708,1989.

17. Rubio, M. P., von Deimling, A., Yandell, D. W., Wiestler, 0. D., Gusella, J. F., andLouis, D. N. Accumulation of wild type p53 protein in human astrocytomas. CancerRes., 53: 3465—3467,1993.

18. Vogelstein, B., and Kinzler, K. W.p53 function and dysfunction. Cell, 70: 523—526,1992.

19. Bigner, D. D., Bigner, S. H., Pontén,J., Westermark, B., Mahaley, M. S., Ruoslahti,E., Herschman, H., Eng, L. F., and Wikstrand, C. J. Heterogeneity of genotypic andphenotypic characteristics of fifteen permanent cell lines derived from human gliomas. J. Neuropathol. Exp. Neurol., 40: 201—229,1981.

20. Casey, G., Lo, H. M., Lopez, M. E., Vogelstein, B., and Stanbridge, E. J. Growthsuppression of human breast cancer cells by the introduction of a wild-type p53 gene.Oncogene, 6: 1791—1797,1991.

21. Cajot, J. F., Anderson, M. J., Lehman, T. A., Shapiro, H., Briggs, A. A., and Stan

bridge, E. J. Growth suppression mediated by transfection of p53 in Hut292DMhuman lung cancer cells expressing endogenous wild-type p53 protein. Cancer Res.,52: 6956-6960, 1992.

22. Lehman, T. A., Bennett, W. P., Metcalf, R. A., Welsh, J. A., Ecker, J., Modali, R. V.,UlIrich, S., Romano, J. W., Appella, E., Tests, J. R., et aL p53 mutations, rasmutations, and p53-heat shock 70 protein complexes in human lung carcinoma celllines. Cancer Res., 51: 4090—4096, 1991.

23. Davidoff, A. M., Pence, J. C., Shorter, N. A., Iglehart, J. D., and Marks, J. R.Expression of p53 in human neuroblastoma- and neuroepithelioma-derived cell lines.Oncogene, 7: 127—133,1992.

24. Saylors, R. B., Sidransky, D., Friedman, H. S., Bigner, S. H., Bigner, D. D., Vogel

stein, B., and Brodeur, G. M. Infrequent p53 gene mutations in medulloblastomas.Cancer Res., 51: 4721—4723, 1991.

25. Lu, X., Park, S. H., Thompson, T. C., and Lane, D. P.Ras-induced hyperplasia occurswith mutation of p53, but activated ras and myc together can induce carcinomawithout p53 mutation. Cell, 70: 153—161,1992.

26. Ionov, Y., Peinado, M. A., Malkhosyan, S., Shibata, D., and Perucho, M. Ubiquitoussomatic mutations in simple repeated sequences reveal a new mechanism for coloniccarcinogenesis. Nature (Lond.), 363: 558—561,1993.

27. Mercer, W. E., Shields, M. T., Amin, M., Sauve, G. J., Appella, E., Romano, J. W., and

Ullrich, S. J. Negative growth regulation in a glioblastoma tumor cell line thatconditionally expresses human wild-type p53. Proc. Natl. Acad. Sci. USA, 87: 6166—6170, 1990.

28. Tabuchi, K., Fukuyama, K., and Mineta, T. Cytogenetic diagnosis of brain tumors.Brain & Nerve, 44: 871—879,1992.

652

nisms exists in the literature (25, 26). Whether these are alternativetransformation pathways or overlapping oncogenic routes where the

knockout of a downstream effector of p53 action replaces absence ofp53 inactivation remains to be established. Finally, because restorationof wild type p53 has been shown to induce growth arrest in T98Gglioblastoma cells (which express endogenous mutant p53 but fail toform tumors in mice) (27) it will be of interest to see whether the sameoccurs in those cell lines that express WTpS3 and thereby to addressthe question ofp53 gene dosage effects, as well as in those cells whichare intrinsically able to form tumors in experimental animals.

Acknowledgments

We thank Dr. D. Bigner for providing the gliosarcoma cell line D247MG,Dr. P. Shaw for oligonucleotide synthesis, and Dr. B. Vogelstein for providingplasmid constructs. We are grateful to Dr. B. Sordat for providing the nu/numice and to Dr. J. Costa for providing the sterile animal facility. We acknowl

edge the expert technical assistance of R. Jaufeerally and M-F. Hamou.

References

1. Russell, D. S., and Rubinstein, L. J. Pathology of tumours of the nervous system.London: Edward Arnold, 1989.

2. Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell,61: 759—767,1990.

3. Cavenee, W. K., Scrable, H. J., and James, C. D. Molecular genetics of human cancerpredisposition and progression. Mutat. Res., 247: 199—202,1991.

4. Louis, D. N., von Deimling, A., Chung, R. Y., Rubio, M. P., Whaley, J. M., Eibl, R.

H.,Ohgaki,H.,Wiestler,0. D.,Thor,A. D.,andSeizinger,B.R.Comparativestudyof p53 gene and protein alterations in human astrocytic tumors. J. Neuropathol. Exp.Neurol., 52: 31—38,1993.

5. Reifenberger,G., Liu, L., Ichimura,K., Schmidt,E. E., andCollins,V. P.Amplification and overexpression of the MDM2gene in a subset of human malignant gliomaswithout p53 mutations. Cancer Res., 53: 2736—2739,1993.

6. Hayashi, Y.,Yamashita, J., and Yamaguchi, K. Timing and role ofp53 gene mutationin the recurrence of glioma. Biochem. Biophys. Res. Commun., 180: 114@—1150,1991.

7. Sidransky, D., Mikkelsen, T., Schwechheimer, K., Rosenblum, M. L., Cavenee, W. K.,and Vogelstein, B. Clonal expansion of p53 mutant cells is associated with braintumour progression. Nature (Lond.), 355: 846—847,1992.

8. Van Meir, E., Ceska, M., Effenberger, F., WaIz, A., Grouzmann, E., Desbaillets, I.,Frei, K., Fontana, A., and de Tribolet, N. Interleukin-8 is produced in neoplastic andinfectious diseases of the human central nervous system. Cancer Res., 52: 4297—4305,1992.

9. Van Meir, E., Sawamura, Y., Diserens, A. C., Hamou, M. F., and de Tribolet, N.Human glioblastoma cells release interleukin 6 in vivo and in vitro. Cancer Rca., 50:6683—6688, 1990.

10. Kikuchi, T, Rempel, S. A., Rutz, H. P.,de Tribolet, N., Mulligan, L, Cavenee, W. K.,Jothy, S., Leduy, L., and Van Meir, E. Turcot's syndrome of glioma and polyposisoccurs in the absence of germ line mutations of exons 5 to 9 of the p53 gene. Cancer

Research. on December 9, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

1994;54:649-652. Cancer Res   Erwin G. Van Meir, Tetsuro Kikuchi, Mitsuhiro Tada, et al.   Glioblastoma Cells

Gene and Its Expression in Humanp53Analysis of the

  Updated version

  http://cancerres.aacrjournals.org/content/54/3/649

Access the most recent version of this article 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. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/54/3/649To request permission to re-use all or part of this article, use this link

Research. on December 9, 2020. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from