Hereditary and acquired p53 gene mutations in childhood acute lymphoblastic leukemia

Hereditary and acquired p53 gene mutations in

childhood acute lymphoblastic leukemia.

C A Felix, … , J J Letterio, J Whang-Peng

J Clin Invest. 1992;89(2):640-647. https://doi.org/10.1172/JCI115630.

The p53 gene was examined in primary lymphoblasts of 25 pediatric patients with acute lymphoblastic leukemia by the RNase protection assay and by single strand conformation polymorphism analysis in 23 of 25 cases. p53 mutations were found to occur, but at a low frequency (4 of 25). While all four mutations were identified by single strand conformation polymorphism, the comparative sensitivity of RNase protection was 50% (2 of 4).

Heterozygosity was retained at mutated codons in 3 of 4 cases. One pedigree was consistent with the Li-Fraumeni syndrome, and bone marrow from both diagnosis and remission indicated a germline G to T transversion at codon 272 (valine to leucine).

Although members of another family were affected with leukemia, a 2-bp deletion in exon 6 was nonhereditary. The other two nonhereditary p53 mutations included a T to G

transversion at codon 270 (phenylalanine to cysteine) and a G to C transversion at codon 248 (arginine to proline). These data support the role of both hereditary and acquired p53 mutations in the pathogenesis and/or progression of some cases of childhood acute lymphoblastic leukemia.

Research Article

Find the latest version:

Hereditary and

Acquired

p53 Gene

Mutations

in

Childhood Acute Lymphoblastic Leukemia

Carolyn A. Felix,* Marion M. Nau,t TakashiTakahashi,*TetsuyaMitsudomi,*ItsuoChiba,t David G. Poplack,*

Gregory H.

Reaman,*

Diane E. Cole,* John J. Letterio,* Jacqueline Whang-Peng,11 Turid Knutsen,11 and John D.Minna"

*Pediatric, WNavy Medical Oncology, and

IlMedicine

Branches, National Cancer Institute, National Institutes ofHealth, Bethesda, Maryland 20892; Division ofHematology-Oncology, Children's National Medical Center, and Department ofPediatrics,

George Washington University School ofMedicine, Washington, District ofColumbia 20010; and 'Simmons Cancer Center, University of Texas Southwestern School ofMedicine, Dallas, Texas 75235

Abstract

The p53 gene was examined in primary lymphoblasts of 25

pediatric patients with acute lymphoblastic leukemia by the

RNase protection assay and by single strand conformation polymorphismanalysisin23 of25cases.p53 mutationswere

foundtooccur,butatalowfrequency(4of25).While allfour

mutationswere identified by singlestrand conformation

poly-morphism,thecomparative sensitivityof RNaseprotectionwas

50%(2 of4).Heterozygositywasretainedatmutated codons in 3 of 4cases.Onepedigreewasconsistent with the Li-Fraumeni syndrome, and bonemarrowfrom bothdiagnosisand remission indicatedagermline Gto Ttransversionatcodon 272(valineto

leucine). Although members of another family wereaffected

with leukemia, a 2-bp deletion in exon6 was nonhereditary.

The othertwononhereditaryp53mutations includedaTtoG

transversionatcodon270

(phenylalanine

to

cysteine)

andaG

toCtransversionatcodon248

(arginine

to

proline).

Thesedata

supportthe role of bothhereditaryandacquired p53mutations

inthepathogenesis and/or progressionofsome casesof

child-hood acute

lymphoblastic

leukemia.

(J.

Clin. Invest. 1992.

89:640-647.) Keywords: B-cellprecursor *germline*

Li-Frau-meni syndrome*T cell* tumor suppressor gene

Introduction

Afterthediscovery ofthe 8;14 translocation and the

conse-quent activation of the c-myc geneinBurkitt'slymphoma(1), studiesofthe molecular basisofchildhood acutelymphoblastic leukemia(ALL)'haveconcentrated uponsearchingfor poten-tial dominant oncogenes at breakpoint regions of

chromo-somal translocations(2).Insolidtumors of both children and

adults,retinoblastomabeingtheparadigm,anadditional focus

hasbeen theidentificationof genesofadifferentclass,which

functionintheirwild-typeform as recessiveoncogenesor

tu-Addressreprintrequests to Dr. Felixather current address: Division of

Oncology, Department ofPediatrics,WoodBuilding,4th Floor,

Chil-dren'sHospitalofPhiladelphia, 34th Street and Civic Center

Boule-vard, Philadelphia,PA 19104.

Receivedfor publication2July1991 and inrevisedform27

Sep-tember1991.

1.Abbreviationsused in thispaper:ALL,acutelymphoblastic

leuke-mia;BM, bone marrow; CML, chronic myelogenous leukemia; ORF,

openreading frame;PB,peripheralblood; PCR, polymerase chain

reac-tion; SSCP, singlestrandconformation polymorphism.

TheJournal of ClinicalInvestigation,Inc.

Volume89, February 1992, 640-647

morsuppressorgenes (3).Onesuchrecessiveoncogene isp53 (4-12), located on chromosome 17 at band p13.1 (13, 14).

Inactivation ofthe human p53 gene,whetherby gross

struc-tural change, homozygous deletion, or point mutation, has

sincebeenimplicatedin thepathogenesisoflung, breast, colon,

brain, and liver cancers, chronic myelogenous leukemia

(CML) and childhood osteogenic and rhabdomyosarcomas

(15-26).Moreover,recentstudieshavedemonstratedthe

pres-enceof germline p53 mutations in Li-Fraumeni pedigrees

af-fected with breast cancers, sarcomas, and braintumors (27,

28). The observation of Li-Fraumeni syndrome-type cancers,

including lymphomas, in transgenic mice overexpressing a

mutantp53hassuggestedapotentialroleofp53 inactivationin

thepathogenesis ofhuman lymphoid malignanciesaswell (29).

Sinceacuteleukemiasareconsideredcomponenttumorsofthe

Li-Fraumenisyndrome,weundertooktostudythe p53 genein

childhoodALL.

Methods

Samplecollection. Materialswerecollectedperprotocol or as part of

standardcare.Lymphoblasts from 80children withBcell precursor

ALL, 21withTcell ALL, and three infants, aged 0-21 yr at diagnosis,

wereobtainedandcharacterizedaspreviously described (30, 31).

Sam-pling timeswere atdiagnosis,atbone marrow (BM) relapse, or at

fail-ure ofinduction chemotherapy. BM collectedduring remissionwas alsoavailableinonecase.Cytogenetic analysiswasby standard meth-ods(32).

DNAandRNApreparationand Southern analysis.High molecular

weightDNA and total cellular RNA were isolated as described (30, 31).

For standard Southernanalysis 10,gofgenomicDNAsdigestedby

BamHI(Bethesda Research Laboratories,Bethesda, MD),EcoRI

(Be-thesdaResearchLaboratories),orHindIII(Boehringer Mannheim

Bio-chemicals, Indianapolis, IN)washybridized witha1.8-kbhumanp53 cDNA-containing XbaI fragment derived fromclonephp53c I(12,15).

Three chromosome 17ppolymorphic regionswereassessed byusing

pBHP53(33), pMCT35.1(34),andpYNZ22(35) fragmentsasprobes

andBamHIorMspI (BethesdaResearchLaboratories)cleaved DNA.

Both thepBHP53andMCT35.1 probeswerepreassociatedwithan

5,000-fold excessofsonicated, alkali-denatured human placental

DNA(SigmaChemical Company, St. Louis, MO) for 2 hbefore hybrid-ization,andnitrocellulose filterswereprehybridizedovernightin a

so-lutioncontaining100

gsg/ml

ofthe same reagent.

Screeningformutations. RNaseprotectionassayswereperformed

aspreviously described usingthreeriboprobes(p53XP, p53PA, p53M)

spanningthep53openreadingframe(ORF;reference15).The

poly-merase chain reaction/single strand conformation polymorphism

(PCR/SSCP)method(36-38),asmodifiedtoscreen forpoint

muta-tionsof thep53gene ingenomicDNA, also has been described (39),

Characterization of abnormalities suggested by screening. RNase protection abnormalitieswere verified by cDNA sequencing.

First-strand cDNA wassynthesized using5 ,ug of total cellular RNA and a

p53-specific primer, followedby PCRamplification oftheentirep53 ORF(15).PCR products were cleaved with EcoRI, agarosegel-purified

by the Geneclean method(Bio101, La Jolla, CA), andligatedinto the

EcoRI site of pGEM-7Zf+ (Promega Corp., Madison, WI) for transfor-mation of DH5a cells (Bethesda Research Laboratories). The entire

p53 ORFof individualcDNA subclones was sequencedusing

Seque-naseVersion 2.0 (U.S.Biochemical Corp., Cleveland, OH), SP6 and

T7sequencing primers, and four sense oligonucleotides which have

beendescribed(15). Mutationswereconfirmedbysequencingthe ap-propriate portionof the ORF in the opposite direction.

Mutationswerealsoconfirmedbyrestrictionenzyme digestionof

cDNAsubclonesifarestriction sitewasaltered and byrestriction

en-zymedigestionordirect sequencing ofgenomic DNA/PCR products. 1

;igofgenomicDNA wasPCR-amplifiedand 1/20 to 1/250 ofthese

products was used as template in a second heminested PCRreaction

forsequencing.30cyclesat95°Cfor 1min,58°Cfor 1min,and72°C

for 3 minwereutilized.PCRproductswereagarosegelisolated,

Gene-clean-purified (Bio 101), and 500 ngwas directly sequenced using

nestedoligonucleotides.

Family studies. Familyand medical histories were obtained by

chartreview orinterview. Peripheral whiteblood cells were used in the

studies ofparents,siblings,and normalindividualsand were collected

after explanation of studiestobeperfoimed. Restrictionenzyme

diges-tion or directsequencing of genomicDNA/PCR products, and SSCP

analyses wereperformedasappropriatetoindividual cases.

Results

Identification

ofpS3mutations by RNase protection and SSCP analysis. Whilethesensitivityof the RNaseprotectionassay is - 50%,it isspecific and avoids PCR artifact (40) andwasthus

employed as a method ofdetecting small mutations. 25

pa-tientsincluding 12 children andtwoinfantsdiagnosed withB

cellprecursor ALLand11 children withTcellALL were stud-ied. Theirleukemiccellswereobtainedatdiagnosis(15of25),

at BM relapse (8 of 25), orboth(1 of25), and inone case at

failureof induction chemotherapy. 2 of 25 childrenwere iden-tified with RNaseprotectionpatternssuggestive of mutations. SSCPanalysiswasperformed in 23 of 25casesand,consistent with the 50% sensitivity ofRNaseprotection, verifiedthose two

mutations and alsoidentifiedtwoothers (Table I). Unlikethe

RNaseprotectionassay,under theexperimental conditions

de-scribed above 90% (36 of 39) of knownpointmutations of the

p53gene aredetectablebySSCP(39).

Absence of obvious rearrangement or deletion of the p53

gene in childhoodALL. Cytogenetic studies of the p53 gene

wereperformed intheleukemiccellsof33patients,including 7

of the 25 screenedbyRNaseprotection, and revealed an ab-sence of obvious structural abnormalities of chromosome 1 7p

and,in one case, a triploid karyotype. Southern analysisofthe p53 gene was performed in 101 casesincluding 22 of the 25

screened by RNase protection and showed only normal pat-terns and no evidence of rearrangement or deletion (not

shown). LymphoblastDNAofasinglechild withBcell precur-sor ALL atdiagnosis manifestedafragmentof alteredsize in EcoRI, butnotineitherBamHIorHindIIIdigests. Although

exons 5-8appeared normal by SSCPanalysis, eitherasmall mutationorpolymorphismin anotherregionofthep53geneis

possibleinthiscase.

Characterization ofap53mutation in a caseofchildhood T

cellALL.RNaseprotectionassayofperipheralblood(PB)

lym-phoblastmRNAofachild withrelapsedTcellALL(T-ALL

Pt.16)revealed relatively abundant p53messageand suggested thepresenceofbothanormal allele andapointmutationon

the other allele whichlocalizedtothe PA fragment inaregion 3'oftheMprobe (not shown). SSCP analysis also suggesteda

mutation which localizedto exon 8(Fig. 1). cDNAsequencing

ofthep53 ORF revealedapoint mutation(TTTtoTGT,

phe-nylalaninetocysteine)atcodon 270, thus confirming the

find-ings ofboth RNaseprotectionandSSCP (Fig. 2). This

muta-tioncreates anovel PvuIIrestrictionsite,andPvuIIdigestion

ofa PCR-amplified fragment oflymphoblast genomic DNA

and of additional cDNA subclones verified the presence of

bothmutantandwild-typealleles (not shown).

Characterization oftwo p53 abnormalities in a case of

childhood B-cellprecursor ALL.RNaseprotectionassayofPB

lymphoblast mRNAofachild with relapsedBcellprecursor

ALL(Pre-BALLPt.4) showeda relatively low level ofp53

mRNA. The absence of mRNAsfullyprotected byeitherthe

PA or Mprobessuggestedtheabsenceofanormal allele.

Pro-tectionoffourfragments bythe 3'PAprobewhosesumlength

exceededthe sizeof complete

protection

suggestedthe

pres-enceofmorethanonemRNA(notshown). Oneabnormality

was presentin theregion ofthe M probeoverlappingPAand

was morespecificallylocalizedto exon6 by SSCP(Fig. 1). The

presenceoftwo mRNA species,both witha2-bpdeletion at

p53 codons214/215 whichwouldcause aframe shiftanduse

ofapremature TGAtermination codon in exon 6 was

con-firmedby cDNAsequencing (Fig. 2).Directsequencing of this

area in PCR-amplified lymphoblast genomic DNA revealed

thatonlythemutantallelewaspresent.

One cDNAsubclone contained not only the codon 214/

215deletion,but alsoa133-bp insertionatthe precise junction

oftheunaltered3'and5'endsofexons9 and 10,respectively

(Fig. 2). This cDNA insertion sequence matched that of

se-TableI.Summary ofMutations inthep53Gene inChildhoodALL

Amino acid RNase

Patient Exon Codon Type Mutation change Zygosity Origin Time protection SSCP

T-ALL16 8 270 tv TTT to TGT phetocys hetero non-hered relapse + +

Pre-B 4 6 214/215 del/fs na na homo non-hered relapse + +

Infant 3 7 248 tv CGG to CCG arg to pro hetero non-hered relapse (notatdx) - +

Pre-B80 8 272 tv GTG toTTG val toleu hetero germline dx(remission) - +

Abbreviations: tv, transversion; del, deletion; fs, frameshift; na, not applicable; hetero, heterozygous; homo, homozygous; non-hered,

Pre-B

ALL

Pt. 4

Z 4

... 't 0

kwok

Infant

ALL

Pt.

3

x ₌

4.C 0

Pre-B

ALL

Pt. 80

.E

I I c& X -G)

7

L .. ..

8

L

wresti.

E4..

*IP

126 187 187 22

AatI1

271 bp 167 bp

!25 261 261 30'

Dra I

I

425 bp 245 bp

Figure 1. Identification and

localization of mutationsto

specificexonsof thep53gene

by the PCR/SSCP method

(39). Patientswerestudiedat

diagnosis, relapse,or

remis-sionasindicated, and DNAs

offamily membersor

indi-vidualswithout p53

muta-tionsservedascontrols. Sites

7 ofrestrictionenzyme

cleav-ageingenomic DNA and

re-4 sultantnormalsizesof

geno-micDNA/PCR fragments

containing individualexons

areshownby schematic.

quencebeginningatnucleotide(nt)196of the 2.5-kb intron9

in normal genomic DNA. The intron 9 genomic DNA se-quenceof thispatient's lymphoblasts wasidenticalto that of

eightnormalindividuals and didnotcontainamutationwhich

would favor alternativesplicing (41)in regions including: (a) the 133alternativelysplicednucleotides(nt196-328 ofintron 9) (see Fig. 2); (b) targetsequences 30 bp upstreamof this

re-gion (taactaac)and upstream ofexon 10(tacttac);and(c) splice

T-ALL

Pt.

16

0

CLL >

5

6

l0

270

21...2-

:::::A.- ...

Jo::

::

~~~~~~~~~~~~~~~~.

... ... ... .-.--:::-:::-.--...:---::.

ATG TTT1-1TGT

Pre-B ALL Pt. 4

_TAdeletion

214/215

TGA

,.,,, 4 ~~~~~~~~..._.. .

.12::.:

:0

31...1

ATG TGA TGA

TA deletion 214/215

2

I

....

....

... ...

.8

.

2 1~~~~~~~~:4: ..2

[3

1

7 8 9 | 10 |1

...

ATG

/~~~~~~~~~~~~~~~~ TGA

gaccagaccagctttcaaaaagaaaa ttgttaaagagagcatgaaaatggtt ctatgactttgcctgatacagatgcta ct(g)cttacgatggtgttacttcctg

|ataaactcgtcgtaagttgaaaatatt

248

CGG -*CCG

Pre-B ALL Pt. 80

---..---...---..---...---...--...---..---...---...--...---...--...;;,;;;...;;...;...-;----;;;--

....-.;.-;;--... .. .. ..

...3

ATG

272

GTG -*TTG

Figure2. Schematic of

ab-normalities in thep53 ORF

identified inlymphoblastsby

cDNA(T-ALLPt.16,Pre-B

ALLPt.4)and/or direct

(Pre-B ALL Pt.4, InfantALL

Pt.3, Pre-B ALL Pt.80)

methodsofsequencing.The

shadedregionsrepresent

ORFsequences.Datawere

consistent with the presence

ofboth normal and mutant

allelesinlymphoblast DNA

in allcasesshown _except

Pre-B ALL Pt.4 whereonly

the mutant allele is present. The lymphoblast genomic

sequencesurroundingthe

in-tron9insertion in Pre-B ALL

Pt. 4is as follows:

5'-ccaact-K

tataccataatatatattttaaagGAC-CAG... .AATATTgtaatgtt-gaaaatggatttaatacaccta-3'.

consensus sequences attheexon9/intron9 boundary (donor: TCAGgtacta) and the intron 9/exon 10 boundary (acceptor:

ctgcagATCC) and sequencesflankingthe 133-bpintron 9 in-sertion. (For genomicsequencessurrounding this insertion,see

legendtoFig. 2.) Thesedatacorroborate all RNase protection

fragmentsidentifiedand areconsistent withp53 allele loss and

low-leveltranscriptionoftworelated, butdifferentmRNA

spe-cies by alternative splicing fromasingle mutantallele.

Characterization ofap53 mutation in a case ofALL of

infancy. Although RNase protection assay of lymphoblast

mRNAofaninfant withBcellprecursor ALL at relapse(Infant

ALL Pt.3) showed full-length protection with all three ribo-probes (not shown), SSCPanalysis ofgenomicDNAfromthe

samesamplingsuggested bothanormal alleleand an abnormal

allelewith amutation in exon 7 (Fig. 1), whichbydirect

se-quencingwasfoundtobe a G toCtransversionatp53 codon

248(CGG to CCG, arginine to proline) (Fig. 2). Incontrast,

repeated SSCP analyses oflymphoblast genomic DNA

sam-pledatdiagnosisrevealed exon7-containingPCRfragments of onlythe normal size, suggesting that the mutation had been

acquiredatsometime duringthe courseoftherapy.

Agermlinep53mutationinchildhoodBcellprecursorALL.

The leukemic cellsofanadolescentmale (Pre-B ALL Pt.80)

showed atriploidkaryotype.Nop53mutation was detected by theRNaseprotection assay (not shown), but SSCP analysis of BM lymphoblast genomicDNAsuggested thatamutationin exon 8 was present at diagnosis (Fig. 1). At least one normal sequence and a G to Ttransversionatp53 codon272(GTGto

TTG, valinetoleucine)werefoundbydirectsequencing(Fig. 2). Thus, heterozygosity appeared to be retained, although somecontribution to the normal sequence by nonleukemic cells is possible. Both SSCP analysis and direct sequencing showed that mutantand normalalleleswerealsopresent in a

remission BM sample where morphological examination showed no evidence of disease (Fig. 1). These data indicate that themutation atp53codon 272 wasgermline and likely heredi-tary(c.f. subsection Familystudies below).

Southernanalysisofpolymorphic regions.Southern analy-sisof thepBHP53, pMCT35.1,andpYNZ22polymorphicloci wasperformedonlymphoblastgenomicDNAs inall 25 cases

studied byRNaseprotectionandin the caseshowinganovel EcoRI site.Homozygouspatterns were found in two ofthe four caseswith definitive mutations and in the case with a novel EcoRIsite.Incontrast,heterozygouspatternswereobserved at one or moreoftheseloci in the other 23 casesincludingtwo

where mutationswere present. These findings support

chro-T-ALL Pt. 16

ATG TGA

Infant

ALL

Pt. 3

... ... ... ...

...X...

... ... ...

---mosome 17p allele loss in one case with evidence of mutation of the

p53

genewhere

only

themutant

p53

allelewasidentified (Pre-B ALL Pt. 4) and verify heterozygosity at chromosome 17p in two cases where both normal and mutant p53 alleles werefound(InfantALL Pt. 3 and Pre-B ALL Pt. 80). In the caseof T-ALL Pt. 16, the leukemic cells showed one normal p53 sequence and an abnormal sequence with a mutation at p53 codon 270, yethomozygouspatternsatall three polymor-phic loci. Based upon the expectedfrequenciesof

heterozygos-ityattheseloci (49%, 38%, and 86%,respectively),thechance that all would behomozygous in the same individual is 4% (33-35). Despite the small chance ofthis combination, and since both normal andmutantalleles were present,these

find-ingsareconsistentwith retention ofbothalleleswhichwere by

chance

homozygous

ateachof the three 17p

polymorphic

loci.

Family studies. Medical and family histories were available

on 18 of the 25patients studied byRNaseprotectionand on thechild whoselymphoblastDNAshowed a novel EcoRI site. The younger brotherofthepatient with a germline mutation at p53 codon 272 was recently diagnosed with osteogenic sar-coma, and the mother ofthepatientandfivematernal grand-parentsortheirsiblingswere affectedatagesasyoungas30 years withlungandother cancers.This kindredmay thus repre-sent aLi-Fraumenifamily (Fig. 3, top). However,in the three

othercasesofchildhoodALLanalysis ofp53sequence in

paren-talgenomicDNAindicatedthat thep53 mutationswere

nonhe-reditary (not shown). Despite

the

acquired

natureof these

mu-tations,thepatientwhoselymphoblastscontainedap53 codon

214/215deletionhad abrotheranddistant cousin with

child-hood acute leukemia and a family history of leukemia and

breast,gastrointestinal,and prostatecancers overfour

genera-tions ofadults(Fig. 3, bottom).Thefamilies of the othertwo

children whose lymphoblasts contained p53 mutations were

affectedbycanceronly during adulthood,and thuswere not

suggestive of the Li-Fraumeni syndrome. Thefamilies of 14

patients

whose

lymphoblast p53

geneswerenormal andof the

patient-with a novel EcoRIrestriction site wereeither

unaf-fected bycancer(10 of15),oraffected only

during

adulthood

(5of15). Inaddition,onechild with normallymphoblast p53

genes had apast

history

of

Ewing's

sarcoma, and another a

historyremoteby 14 yearsofpreviousALL.

Discussion

Theimpetus for this investigation of p53 mutationsin

child-hood ALL, whetherhereditaryoracquired,wasthereported

findingthatmice transgenic fora mutantp53genedeveloped

lymphoid tumors as well as other Li-Fraumeni syndrome

cancers (29). This syndrome of

multiple

primary

cancers, which occur at an early age in either individuals or infamilies,

was firstdescribed in 1969 (42). Softtissue sarcomas, bone,

brain,andbreast cancers, adrenal corticalcarcinomas,and

leu-kemias are considered component tumors of this syndrome in thehuman. In pedigrees with breast cancers, sarcomas, and brain tumors, the presence ofgermline p53 mutations in a

regioncontainingcodons 245, 248, 252,and258 has suggested

that p53 may be the cancersusceptibilitygene(27, 28). The present study demonstrates that in childhoodALL, another component tumor, the p53 genesometimesmay be alteredby

small mutationswhichareeitherhereditary oracquiredand which would lead to changes in predicted p53 peptides (Ta-bleI).

Thelymphoblastsofonechild withrelapsedTcell ALL and

an acquired p53 mutation showed a TtoG transversion at

codon 270 that would result in amino acid substitution

(phenyl-alanineto cysteine) anda change in charge inahighly

con-servedregioninvolved inSV40 largeT antigen binding (5,12).

In another child with B cell precursorALL,whosep53

muta-tionwashereditary,acodon in thissameregion, codon272,

was abnormal (GTG to TTG, valine to leucine). Although these specific mutations have not been observed in other cancers, many mutations clusterin regions of the gene

con-servedin evolution and it has been speculated that theydisrupt

theregulatoryinteraction ofp53withaputativecellular

coun-terpartof large T (15, 17, 18, 19,22).

Thelymphoblasts of another child with relapsed B-cell

pre-cursorALL showedanonhereditaryhomo-orhemizygous

2-bpdeletionatp53 codons 214/215. The resultant frame shift and premature termination inexon6shortens the length ofthe p53 protein product by 44% and eliminates two of the evolu-tionarily conserved domains,oneof the two SV40 large T anti-genbinding sites, and the nuclear localization signal (5, 12, 43). This mutation in exon 6 may also have changed the conforma-tion of the DNA and thus the accessibility of target sequences and/or splice junctions downstream, as a unique mRNAwas

identified which contained both the codon 214/215 deletion and a 133-bp insertion alternatively spliced between exons 9 and 10. This alternatively spliced p53 mRNA is distinct from anypreviously described (16, 17) because it arose by splicinga

region whichwasalready flanked at its 5' end inwild-typeform by the 5'-ag-3' sequence of a splice acceptor and thus began 196 bp internaltothe start of the 2.5-kb intron 9. No mutation in intron 9 favoring alternative splicing was found. The intron 9 sequenceatthe 3'insertion/intron boundary (5'-gtaagt-3')may be a better splice donorconsensusthanthatnormallyfound at theexon9/intron 9boundary (5'-gtacta-3') (41).This 5'-gtaagt-3' may contain acommonpolymorphismwhich results in al-ternative splicing in that itwasalso found in genomic DNA from several normal individuals. In the leukemic cells witha more5' terminationcodon inexon6 theinsertionin the p53 ORF isinconsequential (Fig. 2). However,thesameinsertion containsaninframe TGA 82bpdownstream that would

elimi-nate exons 10 and 11 from thereadingframe and 62 amino acids from thep53protein if present alone. The alternatively spliced mRNA may also reflectincomplete processingoran undeterminedabnormalityintranscriptionalregulation.

Despitetheearlyage ofaninfant with B cell precursor ALL,

thep53codon 248 mutation observedatrelapsewasalso

non-hereditary and,moreover,presumablyacquiredat sometime

during the course ofor perhapsas aresult oftherapy. This

mutation wouldcause anamino acidsubstitution (arginineto

proline)at aresidue of theproteinknowntobe abnormal in

someindividuals with theLi-Fraumeni syndrome.This residue

again falls in aconserved regionofp53 involved in murine

SV40 largeTantigen binding (5, 12).Therefractorynatureof

the disease ofthis infantatrelapsemayattest tothe importance ofaputativeregulatoryfunction(15, 17-19,22).

Findingsof loss ofheterozygosityare often suggestive of

mutations intumorsuppressor genes.Intheheritable form of

retinoblastoma,forexample,asmallmutation inonealleleis

usuallygermlineandprecedestheloss of thesecondallele in

tumorcells(3,44).Consistent witharecessive

model,

in theB

-II(1

IL

Lung -50y

40y

0

0

Lung

-O5y

Bone

33y

37y

Uterus Dx 35y

20y ALL 17y

Dx17y Osteogenic

Death19y Sarcoma

p53Codon 272

GTG-.TTG

26y 24y Acute ALL Leukemia Dx5y

Dx5y Death By

DeathSy p53 Codon 2141215 del.

Figure3.Pedigreesofcancerpronefamilies of children with ALL andgermline (top)ornonhereditary(bottom) p53mutations.(o, o) livingmale

orfemale;(i,*) deceased; (\)affectedwithcancer;(U)cancer,typeunknown.

19, 45-49), there is also clear evidence ofloss of the normal allele.In contrast,childhoodALLsometimesmaydifferfrom

othercancerswherep53 behavesas aclassictumorsuppressor

gene since thelymphoblasts of three other children retained

heterozygosity. Such cases instead suggest the postulated "trans-dominant negative" mechanism, where both alleles

-W

X

LProstate

-50y

U

40y

0

u u n=5 n=4

)

)

_L

u u

4

produce proteins, butmutationof a single allelecontributesto

transformation because thereissequestrationandinactivation

of wild-type protein by the mutantform(50, 5 1).

By two independent screening tests, this study indicates that thefrequencyof p53mutationsinchildhoodALL is rela-tively low (4 of 25). Thislow frequency is in contrastto muta-tions reported in 5 of 10 T cell ALL celllineswhere in vitro selection has been suggested (52).However,this lowfrequency

ofdetectable mutationsdoes not precludeinvolvementof p53

by othermechanisms including transcriptional inactivationor

posttranslational modification.

We also searchedfor thepresence ofgermline p53 muta-tions andexamined medical andfamily historiesfor cancers

characteristicof theLi-Fraumeni syndrome(42). The

Li-Frau-meni family which weidentifieddiffers from others reported to

have hereditary p53 mutations in both the range of tumors

which occurred(childhoodALL,osteogenicsarcoma, and lung cancer) and the codon 272 location of thegermlinep53 muta-tion. Thispoint mutation isalso incontrasttomajor structural

rearrangementsof the p53 gene which have beendescribedin

osteogenicsarcoma(21). These findings predictthatadditional

heterogeneity in tumor types and specific mutations will be

foundasadditional Li-Fraumeni familiesareinvestigated.

The pedigree of another child withBcell precursor ALL

also seems consistent withacancer-prone

family, having

two other cases ofchildhood leukemia and leukemia and other cancersoverfour generations ofadults. The

p53

codon214/

215deletion which wasfound,

nonetheless,

wasnonhereditary

and occurred inaregion ofthe genedistinct fromthat involved

inpreviously reported Li-Fraumeni families.

Moreover,

intwo

otherpatientswhere ALLrepresentedasecond cancer,nop53

mutationwasidentified atall.These data suggest thata p53

mutationmay notalwaysbeinheritedor evenpresentin

cer-tain cancer-prone individuals and that a mutantgene other

thanp53may beinheritedinfamilieswith

multiple

members affectedby leukemia.

Thus,

inchildhoodALLvariantcancer

susceptibility

syndromes anda more multifactorial

pathoge-neticprocessarepossible.

Acknowledaments

WewishtothankA.Chauvenet and Y. Ravindranath for patient

re-ferrals;J. Fedorko and W.Goldschmidtsforpreparation

ofoligonucleo-tides; P. Chumakovforproviding the entire sequence of thehuman

p53 gene; and F. Li, D. D'Amico,L.Goldstein,D.Jones,and K.Bhatia

for helpful discussions.

CarolynA.Felix isrecipient of the American Society ofPediatric

Hematology-OncologyYoungInvestigator Award(1990).

References

1.Battey, J., C. Moulding, R. Taub, W. Murphy, T. Stewart, H. Potter, G. Lenoir, and P. Leder. 1983. The human c-myc oncogene: structural conse-quences of translocation into the IgH locus inBurkitt lymphoma.Cell. 34:779-787.

2. Testa, J. 1990. Chromosome translocations in human cancer.CellGrowth

Differ. 1:97-101.

3. Ponder, B. 1988. Gene losses in humantumours.Nature(Lond.). 335:400-402.

4. Lane, D., and S. Benchimol. 1990. p53: oncogene oranti-oncogene?Genes Dev. 4:1-8.

5. Soussi, T., C. Caron de Fromentel, and P. May. 1990. Structural aspects of the p53 protein in relation to gene evolution. Oncogene. 5:945-952.

6. Lamb, P., and L. Crawford. 1986.Characterizationof the human p53 gene.

Mol.Cell. Biol.6:1379-1385.

7.Buchman, V.,P.Chumakov,N.Ninkina,0.Samarina,and G.Georgiev.

1988. A variation in thestructureoftheprotein-coding regionofthe humanp53

gene.Gene. 70:245-252.

8.Finlay,C.,P.Hinds,and A. J. Levine. 1989. Thep53 protooncogenecan

actas asuppressoroftransformation.Cell.57:1083-1093.

9.Eliyahu,D.,D.Michalovitz,S.Eliyahu,0.Pinhasi-Kimhi,and M. Oren. 1989. Plasmids_encodingwildtypep53caninhibitoncogene-mediated transfor-mation. Proc.Natl.Acad. Sci. USA.86:8763-8767.

10.Parada, L.,H.Land,R.Weinberg,D.Wolf,and V. Rotter. 1984.

Coopera-tion betweengeneencodingp53tumourantigenandrasin cellular transforma-tion.Nature(Lond.).312:649-651.

1 1.Chen, P.-L.,Y.Chen,R.Bookstein,andW.-H. Lee. 1990. Genetic

mecha-nismsoftumor suppression bythehuman p53 gene. Science(Wash. DC).

250:1576-1580.

12.Zakut-Houri,R.,B.Bienz-Tadmaor,D.Givol,andM. Oren. 1985.

Hu-man_p53cellulartumor_antigen:cDNA sequence andexpressioninCOS cells.

EMBO(Eur.Mol.Biol.Organ.)J.4:1251-1255.

13.McBride,O.,D.Merry,and D. Givol. 1986. The gene for humanp53

cellular tumorantigenis locatedonchromosome 17 shortarm(17pI3).Proc. Natd.Acad.Sci. USA. 83:130-134.

14.Isobe, M.,B. S.Emanuel,D.Givol,M.Oren,and C. M. Croce. 1986.

Localizationof gene for humanp53 tumourantigentoband 17pl3.Nature

(Lond.).320:84-85.

15.Takahashi,T.,M.Nau,I.Chiba,M.Birrer,R.Rosenberg,M.Vinocour,

M.Levitt,H.Pass,A.Gazdar,andJ. Minna. 1989.p53:afrequenttargetfor geneticabnormalitiesinlungcancer.Science(Wash. DC).246:491-494.

16.Takahashi, T.,D.D'Amico,I.Chiba,D.Buchhagen,and J. Minna. 1990.

Identificationofintronicpointmutationsasanalternative mechanism forp53

inactivationinlungcancer.J.Clin.Invest. 86:363-369.

17.Chiba, I.,T.Takahashi,M. M.Nau,D.D'Amico,D. T.Curiel,T.

Mitsu-domi,D. L.Buchhagen,D.Carbone,S.Piantadosi,H.Koga,etal.(FortheLung CancerStudy Group).1990. Mutations in thep53genearefrequentinprimary,

resected non-small celllungcancer.Oncogene.5:1603-1610.

18.Nigro,J.,S.Baker,A.Preisinger,J.Jessup,R.Hostetter,K.Cleary,S.

Bigner,N.Davidson,S.Baylin,P.Devilee,etal. 1989.Mutationsin thep53gene

occurindiverse humantumourtypes. Nature(Lond.).342:705-708.

19.Baker, S.,E.Fearon,J.Nigro,S.Hamilton,A.Preisinger,J.Jessup,P.

vanTuinen,D.Ledbetter,D.Barker,Y.Nakamura,etal.1989. Chromosome17

deletionsandp53gene mutations incolorectal carcinomas. Science (Wash._DC).

244:217-221.

20.Baker, S.,S.Markowitz,E.Fearon,J.Willson,and B.Vogelstein. 1990.

Suppressionofhuman colorectal carcinoma cellgrowthbywild-type p53.Science

(Wash. DC).249:912-915.

21.Masuda, H.,C.Miller,H. Koeffler,H.Battifora,andM. Cline. 1987. Rearrangementofthep53gene inhumanosteogenicsarcomas.Proc. Nati. Acad.

Sci. USA.84:7716-7719.

22.Mulligan, L.,G.Matlashewski,H.Scrable,andW. Cavanee. 1990.

Mecha-nismsofp53loss inhumansarcomas.Proc.Nati.Acad. Sci. USA.87:5863-5867.

23.Ahujia, H.,M.Bar-Eli,S.Advani,S.Benchimol,andM. Cline. 1989.

Alterations in thep53gene and the clonal evolution of the blastcrisis of chronic myelocyticleukemia.Proc.Nati.Acad.Sci.USA. 86:6783-6787.

24.Mashal, R.,M.Shtalrid,M.Talpaz,H.Kantarjian,L.Smith,J.Beran,A.

Cork,J.Trujilo,J.Gutterman,andA.Deisseroth. 1990.Rearrangementand

expressionofp53in chronicphaseand blast crisis ofchronicmyelogenous

leuke-mia.Blood. 75:180-189.

25.Hsu, I.,R.Metcalf,T.Sun,J.Welsh,N.Wang,andC. Harris. 1991.

Mutationalhotspotin thep53gene in humanhepatocellularcarcinomas. Nature

(Lond.).350:427-428.

26.Bressac, B.,M.Kew,J.Wands,and M. Ozturk. 1991. Selective G to T

mutations ofp53gene inhepatocellularcarcinoma from southern Africa. Nature

(Lond.).350:429-431.

27.Malkin, D.,F.P.Li,L. C.Strong,J. J.Fraumeni,C. E.Nelson,D. H.Kim,

J.Kassel,M. A.Gryka,F. Z.Bischoff,M. A.Tainsky,etal. 1990. Germ linep53

mutations inafamilialsyndromeof breast cancer, sarcomas, and other

neo-plasms.Science(Wash. DC).250:1233-1238.

28.Srivastava, S.,A.Zou,K.Pirollo,S.Blattner,andE.Chang.1990.

Germ-linetransmission ofamutatedp53gene inacancer-prone family with

Li-Frau-menisyndrome. Nature (Lond.). 348:747-749.

29.Lavigueur, A.,V.Maltby,D.Mock,J.Rossant,T. Pawson, and A. Bern-stein. 1989. High incidence of lung, bone, andlymphoidtumorsin transgenic

miceoverexpressingmutantalleles ofthep53oncogene.Mol. Cell. Biol.

9:3982-3991.

30.Felix,C.A.,J. J.Wright, D. G.Poplack,G. H. Reaman, D. Cole,P.

Goldman,andS. J.Korsmeyer.1987. T cell receptor a-, b-, andg-genesin T cell andpre-Bcell acutelymphoblasticleukemia. J. Clin. Invest.80:545-556.

31.Felix,C.A.,D. G.Poplack,G. H. Reaman, S. M. Steinberg, D. E.Cole,

B.J.Taylor,C.G.Begley,andI. R. Kirsch. 1990.Characterizationof

immuno-globulinand T-cell receptorgene patterns in B-cell precursor acutelymphoblastic

leukemia of childhood. J. Clin. Oncol. 8:431-442.

cells without priorinvitro cultureorin vivocolchicine administration. Stain

Tech. 37:17-20.

33.Hoyheim,B., Y. Nakamura, and R. White. 1989. A

BamHI-polymor-phism is detected byagenomic p53-clone (pBHP53).NucleicAcids Res.21:8898.

34. Carlson, M., Y. Nakamura, R. Payson, P.O'Connell, M.Leppert, G. Lathrop, J. Lalouel, and R. White. 1988.Isolationandmapping ofapolymorphic

DNAsequence pMCT35.1 on chromosome 17p[D17S31]. Nucleic Acids Res. 16:783.

35. Nakamura,Y., L. Ballard, M. Leppert, P.O'Connell,G. Lathrop, J.-M. Lalouel, and R. White. 1988.Isolationandmapping of apolymorphicDNA sequence(pYNZ22)onchromosome 17p[D17S30].NucleicAcidsRes.16:5707.

36.Orita,M., H. Iwahana, H. Kanazawa, K.Hayashi,and T.Sekiya. 1989.

Detection of polymorphisms ofhuman DNA gelelectrophoresisassingle strand

conformationpolymorphisms.Proc.Natl.Acad. Sci. USA. 86:2766-2770.

37.Orita, S., Y.Suzuki,T.Sekiya,and K. Hayashi. 1989.Rapidandsensitive

detection of point mutationsand DNApolymorphisms usingthepolymerase

chain reaction.Genomics.5:874-879.

38.Suzuki,Y., M. Orita,M.Shiraishi, K. Hayashi,and T.Sekiya. 1990.

Detection ofras genemutationsin human lung cancers bysinglestrand

confor-mation polymorphism analysis ofpolymerasechainreactionproducts.

Onco-gene.5:1037-1043.

39.Mitsudomi,T., S. M.Steinberg,M. M.Nau, D. Carbone, D.D'Amico,S.

Bodner, H. K.Oie,I. Linnoila,J.S.Mulshine,J. D.Minna,et al. 1991. p53 gene

mutationsinnon-small cell lung cancer cell lines and theircorrelation withthe presenceofrasmutationsandclinical features.Oncogene. In press.

40. Winter, E., F. Yamamoto, C. Almoguera,and M. Perucho. 1985. A

method to detect andcharacterize pointmutations intranscribedgenes:

amplifi-cationandoverexpression ofthe mutant c-Ki-ras allele in human tumor cells.

Proc.Natl.Acad. Sci. USA.82:7575-7579.

41.Mount,S. 1982.Acatalogue ofsplicejunctionsequences. NucleicAcids

Res. 10.

42.Li,F., and J.Fraumeni. 1969. Rhabdomyosarcomainchildren.J.Natl.

CancerInst.43:1365-1373.

43.Addison,C., J. Jenkins, and H.-W. Sturzbecher. 1990. Thep53nuclear

localisation signal isstructurallylinkedto a p34cdc2 kinase motif. Oncogene.

5:423-426.

44.Cavenee, W., T. Dryja, R. Phillips, W. Benedict, R. Godbout, B. Gallie, A. Murphree, L. Strong, and R. White. 1983. Expression of recessive alleles by chromosomal mechanisms in retinoblastoma. Nature(Lond.).305:779-784.

45.Cattoretti,G., F. Rilke, S. Andreola, L. D'Amato, and D. Delia. 1988.p53

expressionin breast cancer. Int.J.Cancer. 41:178-83.

46. Weston, A., J. Willey, R. Modali, H. Sugimura, E. McDowell, J. Resau, B. Light, A. Haugen, D. Mann, B. Trump, et al. 1989.DifferentialDNAsequence

deletionsfrom chromosome 3, 11, 13 and 17 insquamous-cell, large-cell

carci-noma, andadenocarcinoma ofthe human lung. Proc. NatL. Acad. Sci. USA.

86:5099-5103.

47.D'Amico,D., D.Carbone, T.Mitsudomi,M.Nau, J. Fedorko, E. Russell,

B.Johnson, D. Buchhagen, S. Bodner, A. Phelps, et al. 1991. High frequency of

somatically acquiredp53 mutations in small cell lung cancer celllinesand

tu-mors.Oncogene. In press.

48.Varley, J. M., W. J.Brammer,D. P.Lane, J. E.Swallow,C. Dolan, and R. A.Walker. 1991. Lossofchromosome17p13 sequences and mutation of p53 in human breast carcinomas. Oncogene. 6:413-421.

49. Yokota, J., M. Wada, Y.Shimosato,M.Terada, and T.Sugimura. 1987. Loss of heterozygosity on chromosomes 3, 13, 17 in small cell carcinoma and on chromosome 3 in adenocarcinoma ofthelung. Proc.NatL. Acad.Sci. USA.

84:9252-9256.

50.Eliyahu,D., N.Goldfinger,0.Pinhasi-Kimhi,Y.Skurnik, N. Arai, V.

Rotter, and M. Oren. 1988. Meth A fibrosarcoma cells express two transforming mutant p53 species. Oncogene. 3:313-321.

51.Hinds,P.,C. Finlay, and A. Levine. 1989.Mutationisrequiredtoactivate

thep53genefor cooperationwith the ras oncogene andtransformation.J.Virol.

63:739-746.