Identification of mutations in the coding sequence of the proto oncogene c kit in a human mast cell leukemia cell line causing ligand independent activation of c kit product

Identification of mutations in the coding

sequence of the proto-oncogene c-kit in a

human mast cell leukemia cell line causing

ligand-independent activation of c-kit product.

T Furitsu, … , L K Ashman, Y Kanayama

J Clin Invest.

1993;

92(4)

:1736-1744.

https://doi.org/10.1172/JCI116761

.

The c-kit proto-oncogene encodes a receptor tyrosine kinase. Binding of c-kit ligand, stem

cell factor (SCF) to c-kit receptor (c-kitR) is known to activate c-kitR tyrosine kinase, thereby

leading to autophosphorylation of c-kitR on tyrosine and to association of c-kitR with

substrates such as phosphatidylinositol 3-kinase (PI3K). In a human mast cell leukemia cell

line HMC-1, c-kitR was found to be constitutively phosphorylated on tyrosine, activated, and

associated with PI3K without the addition of SCF. The expression of SCF mRNA transcript

in HMC-1 cells was not detectable by means of PCR after reverse transcription (RT-PCR)

analysis, suggesting that the constitutive activation of c-kitR was ligand independent.

Sequencing of whole coding region of c-kit cDNA revealed that c-kit genes of HMC-1 cells

were composed of a normal, wild-type allele and a mutant allele with two point mutations

resulting in intracellular amino acid substitutions of Gly-560 for Val and Val-816 for Asp.

Amino acid sequences in the regions of the two mutations are completely conserved in all of

mouse, rat, and human c-kit. In order to determine the causal role of these mutations in the

constitutive activation, murine c-kit mutants encoding Gly-559 and/or Val-814,

corresponding to human Gly-560 and/or Val-816, were constructed by site-directed

mutagenesis and expressed in a human embryonic kidney cell line, 293T cells. In the

transfected cells, both c-kitR […]

Research Article

Find the latest version:

Identification

of Mutations in the

Coding Sequence of the

Proto-oncogene c-kit

in

a

Human Mast

Cell Leukemia Cell Line

Causing Ligand-independent Activation of c-kit Product

Takuma Furitsu,*Tohru Tsujimura,t Toshiharu Tono,t Hirokazu Ikeda, * Hitoshi Kitayama, *

Uichi

Koshimizu,*

Hiroyuki Sugahara,* JosephH.Butterfield,11 LeonieK.

Ashman,"

Yoshio Kanayama, * Yuji Matsuzawa,*Yukihiko Kitamura,t and Yuzuru Kanakura *

*Second Department of Internal Medicine andtDepartmentof Pathology, OsakaUniversityMedical School, Suita565, Japan;

§ResearchInstitutefor MicrobialDiseases, Osaka University, Suita 565, Japan;I1Departmentof Internal Medicine, MayoClinic, Rochester, Minnesota;and 'Leukaemia Research Unit, HansonCentrefor Cancer Research, Adelaide, South Australia

Introduction

Thec-kit proto-oncogene encodes a receptor tyrosine kinase. Binding ofc-kit ligand,stem cellfactor(SCF)toc-kitreceptor

(c-kitR) is known toactivatec-kitR tyrosinekinase, thereby

leading to autophosphorylation of c-kitR on tyrosine and to

association of c-kitRwith substrates such as phosphatidylino-sitol 3-kinase(P13K).In a human mast cellleukemiacellline

HMC-1,c-kitRwasfoundto beconstitutivelyphosphorylated ontyrosine, activated, andassociated with PI3K without the

addition of SCF.Theexpression of SCFmRNAtranscriptin

HMC-1cells was not detectable by meansof PCR afterreverse

transcription

(RT-PCR)

analysis, suggestingthatthe constitu-tive activation of c-kitRwasligand independent. Sequencingof whole coding region of c-kit cDNA revealed that c-kit genes of HMC-1 cells were composedofa normal,wild-typeallele and a mutant allele with two point mutations resulting in intracellular

amino acidsubstitutions ofGly-560 forVal and Val-816 for

Asp.Amino acidsequences in theregions ofthe twomutations arecompletely conserved in all of mouse, rat, and humanc-kit.

Inorder todeterminethe causal roleofthesemutationsin the

constitutiveactivation, murine c-kitmutantsencodingGly-559 and/or

Val-814,

correspondingtohumanGly-560 and/or

Val-816, were constructed by site-directed mutagenesis and ex-pressed in a humanembryonic kidney cell line, 293T cells. In the transfected cells, both c-kitR (Gly-559, Val-814) and c-kitR (Val-814) were abundantly phosphorylated on tyrosine and activatedin immune complex kinase reaction in the

ab-senceofSCF,whereastyrosinephosphorylation and activation

of c-kitR(Gly-559) orwild-type c-kitRwas modest or little,

respectively.These results suggest that conversion ofAsp-816

toVal in humanc-kitRmay be anactivating mutation and

re-sponsible forthe constitutive activation of c-kitRin HMC-1 cells.(J. Clin.Invest.1993.

92:1736-1744.)

Keywords: proto-oncogene c-kit-point mutation*leukemia-tyrosine kinase-activation

Thiswork was presented in part as a poster at the 34th Annual Meeting

of the American Society of Hematology, Anaheim, CA, December

1992, and was published in abstract form (1992.Blood.80:302a). Addresscorrespondence to Dr. Yuzuru Kanakura, Second Depart-mentofInternal Medicine, Osaka University Medical School, 2-2, Ya-mada-oka, Suita 565, Japan.

Receivedfor publication 23 March 1993 and inrevisedformI June 1993.

The proto-oncogenec-kitwas identifiedasthe cellular homo-logueofthe oncogenev-kit present in the genome of

Hardy-Zuckerman 4-feline sarcoma virus that induces multicentric

fibrosarcomas in the domestic cat ( 1). The proto-oncogene c-kit encodes a transmembranetyrosinekinase receptor that is amemberof receptors for PDGF and CSF- 1 (2, 3).The

char-acteristicfeatures of the receptortyrosinekinasesubfamilyare anextracellulardomain made up of fiveimmunoglobulin-like

repeats, and anintracellular tyrosine kinase,which issplitinto twodomains by ahydrophilic insert sequence( 1-4). The

li-gandfor c-kitproducthasrecentlybeen cloned andvariously designatedas stem cell factor

(SCF),'

mastcellgrowth factor, kit ligand,orsteelfactor(5-12).

Thec-kit gene is allelicwith the dominant whitespotting

(W) locus on mousechromosome5(13, 14),whereas the gene

encoding SCF resides atsteel(SI) locus on mouse chromo-some 10(7,9, 12).Mutationsateither theW/c-kitlocusorthe

SI /SCFlocusaffect variousaspectsof

hematopoiesis

andthe

proliferation and/or

migration

of

primordial

germ cellsand

melanoblasts during embryogenesis (4, 15-17). In

hema-topoiesis, both

WI

c-kit and SI/SCF mutationsaffect the

he-matopoieticstemcellcompartment andgive riseto hypoplas-tic anemia and a deletion ofmast cells (4, 15-17). Recent

experiments

have shown that c-kit product is

expressed

on

mastcells (4, 16, 18) andon asmall,

progenitor-cell

fraction ( - 4%)of adult bone marrow cells(19-21),and that SCFacts

synergistically

with other

hematopoietic

growthfactorsto stim-ulate thegrowth ofa

variety

of

progenitor

cells

(5,

6, 22-24).

Inaddition tonormal hematopoietic cells,weand others

haveinvestigated theexpression and function of c-kit product in human leukemia cells (25-32). In human leukemia cell lines,theexpression of

c-kit

productwasdetected in erythro-leukemiacelllines (HEL, LAMA,and TF-1),

megakaryocytic

leukemiacelllines

(MO7E

andMEG-0 1),andmast-cell leuke-mia cell line

(HMC-l

) (25, 27, 31, 32). Also, c-kit productwas

foundtobeexpressed inmost casesofhuman acute myeloblas-tic leukemia (AML)andinsomecasesofchronic myelogenous leukemia (CML) inblastic crisis(26, 28-31). Further, SCF

inducedrapid tyrosine phosphorylationandactivation of c-kit

product,leadingtoproliferation ofM07Ecells and blast cells in a substantial fraction ofAML cases (27, 28). These results

1.Abbreviations used in thispaper: AML, acute myeloblastic leukemia; CML,chronicmyelogenousleukemia; HMC-1,human mast-cell leu-kemia cell

line-i;

P13K,phosphatidylinositol 3-kinase; rh,

recombi-nanthuman; RTK,receptortyrosine kinase; SI,steel locus;SCF,stem

cellfactor; W, whitespottinglocus.

1736 Furitsuetal.

Abstract

J.Clin. Invest.

©TheAmerican Society forClinical Investigation,Inc.

0021-9738/93/10/1736/09 $2.00

suggestthat c-kit activationmaybeinvolved in some aspectof

excessive proliferation and aberrant differentiation of human leukemia cells. However, structuralabnormalities of c-kitgene contributingtoconstitutive, ligand-independent activation of the c-kitR kinase havenotbeen reported.

In this study, we have investigated the state of tyrosine phosphorylation and activation of c-kit product in human leu-kemia cells, and foundthat,regardlessof stimulationwith SCF, c-kit productisconstitutivelyphosphorylatedontyrosine resi-duesinHMC-1.Neither SCFmRNAtranscript nor biological activity of SCF were detectable fromHMC-1 cells, suggesting

that the activation of c-kit was not due toautocrine mecha-nism. The sequence of thec-kitgene revealed unique, as-yet-unidentified,two-point mutationsin thecytoplasmic domain,

resulting in aminoacid substitutions ofGly forValin codon 560and Valfor Aspincodon 816.Further,introduction ofthe mutated c-kit genes into cells ofa humanembryonic kidney

cellline, 293T, resultedinthegenerationof thec-kit-encoded

protein that wasconstitutively phosphorylatedontyrosineand activated without the addition of SCF.Thus,ourdata provide

the firstevidence that pointmutation(s)inc-kit geneis

capa-ble ofinducing constitutive activation ofc-kit product, and

raisethepossibility that the activating mutation(s) described

here may be involvedinoncogenesisofsomecell types, includ-ing mast cells andhematopoieticstemcells.

Methods

Reagents.Highly purified recombinant human (rh) GM-CSF

(rhGM-CSF)wasagiftfrom Drs. S. Clark and G. Wong(Genetic Institute, Cambridge,MA). rhSCFwasgenerouslyprovided byDr.K.M.Zsebo (Amgen Inc., Thousand Oaks, CA). Theanti-phosphotyrosine anti-body,amurine mAb generatedagainst phosphotyramine,was

gener-ouslysuppliedby Dr. B.Druker (Dana-Farber CancerInstitute,

Bos-ton, MA). Thespecificity of theanti-phosphotyrosine mAbwas de-scribedpreviously(33,34). Rabbit anti-human c-kitserum waskindly

provided by Dr.A.Ullrich (Max-Planck Institute ftirBiochemie,

Mar-tinsried, Germany).This antiserumwasgeneratedagainstasynthetic

peptidecorrespondingtoCOOH-terminal 16 amino acid residues of c-kitR (2).AmurinemAb(YB5.B8) thatreacts with theextracellular domain of human c-kit protein wasoriginally raised against human

AMLblast cells(21,35).Therabbit antiserum againstthe whole

mu-rinec-kit protein (36)wasgenerously donatedby Dr. P.Besmer

(Cor-nellUniversity GraduateSchool of Medical Science, New York). A rat

mAb(ACK-2) againsttheextracellulardomain of murine c-kit prod-uct(19) and the full length of murine c-kit cDNA clone were

gener-ously donatedbyDr.S-I.Nishikawa (Kumamoto University,

Kuma-moto, Japan). Rabbit anti-phosphatidylinositol 3-kinase (PI3K)

serum waspurchasedfrom Upstate Biotechnology,Inc. (Lake Placid,

NY); the

anti-PI3K

serum recognizes an 85-kD subunit ofP13K.

Chemically defined serum-free medium (ASF-102) was purchased

fromAjinomoto(Tokyo, Japan); it containshumantransferrin,

insu-lin,andBSA.

Cell lines. HMC-1, a human mastcellleukemia cell line, estab-lished fromtheperipheral bloodfromapatientwithmastcell leuke-mia, was grown in Iscove's modified Dulbecco's medium

supple-mentedwith 10% horseserum(FlowLab,NorthRyde,Australia)as describedpreviously (37). M07E,ahumanGM-CSF and IL-3

depen-dentcellline,wasobtainedfromDr. SteveClark,GeneticsInstitute,

and wasoriginally established byAvanzi et al. from the peripheral

bloodfromaninfantwithacutemegakaryocyticleukemia(38).M07E

cellswerecultured inRPMI-1640medium (Nacalai Tesque, Kyoto, Japan) supplementedwith 10%FCS(FlowLab)and10ng/ml rhGM-CSF. HEL, a human erythroleukemia cell line, and NTI-4, human

fetus-derived fibroblast cell line, were obtained from the Japanese

Cancer Research ResourcesBank,andthis cell linewere_adaptedto grow and maintained inRPMI 1640 supplemented with 10% FCS.A

humanembryonickidneycellline,293Twaskindly provided byDr. D.Baltimore (TheRockefellerUniversity, New York), and was de-rived fromhumanembryonickidneycellstransformedbyDNAfrom

humanadenovirus type 5(39).293T cellsweremaintained inDMEM

(NacalaiTesque) supplementedwith10%FCS.Inapreliminarystudy,

293T cellswerefoundtobecapableofexpressingagreateramountof c-kitproductthanCOS cellsaftertransfection of c-kit gene.

Flowcvtometry.Cells (5 x

105)

werewashedwith PBS and

resus-pended in PBS containing 0.5% BSA (Sigma Chemical, St. Louis, MO),0.1% NaN3, and human Ig (1 mg/ml)toblockbindingof anti-bodiestoFc receptorsonthe cells.Cellswereincubated with YB5.B8

ormousemonoclonalIgG,control (Becton Dickinson andCo., Moun-tain View, CA)at40Cfor 30min,thenwashed three times with the buffer. Cellsweresubsequentlyincubated withfluorescein-conjugated

goat(Fab')2anti-mouseIgGantibody (Tago, Inc., Burlingame, CA)at

40C for 30min,andwashedtwicebefore analysis using a FACScan® flow cytometer (Becton Dickinson and Co.).

Immunoblotting. The proceduresof cell lysis, gel electrophoresis,

and immunoblotting were performed according tothe methods de-scribedpreviously (27, 28, 33, 34). Briefly, exponentiallygrowing cells

werewashedfreeofserum andgrowthfactors and incubated in

serum-free ASF- 102 medium for 16hat370Ctofactor-deprivethecells. The cells (107 cells suspended in I ml of ASF-102 medium) were then

exposedtorhSCF(100ng/ml)for 15minat370C.Afterstimulation,

cellswere washed with cold PBS and lysedin NP-40 buffer (20mM

Tris-HCl pH8.0, 137mMNaCl, 10% glycerol, 1%NP-40)containing

protease andphosphatase inhibitorsat4°Cfor 20 min.In some

experi-ments, lysisbuffercontaining 1%digitonin (Wako Pure Chemical

In-dustries, Ltd., Osaka, Japan)instead of1% NP-40wasusedtoexamine

theassociation ofc-kit product with P13K. Insolublematerial was

re-moved bycentrifugationat10,000 g for15 min at4°C,and celllysates

weresubjectedtoSDS-PAGE. Proteins were electrophoretically

trans-ferred from thegel onto apolyvinylidenedifluoride membrane

(Im-mobilon; Millipore Corp., Bedford, MA), and immunoblotting was

performedwith eitheranti-phosphotyrosinemAboranti-c-kit serum.

For immunoprecipitation, the lysates from rhSCF-stimulated or unstimulated cells were precleared with protein A-Sepharose beads

(Pharmacia AB, Uppsala, Sweden) for 2 h at4°C. The precleared

ly-sates werethen incubated withaYB5.B8 mAb andprotein

A-Sephar-osebeadstocollect theantigen-antibodycomplexes.The immunopre-cipitates werewashed fivetimes with lysisbuffer containing protease andphosphataseinhibitors,andsubjectedtoSDS-PAGEand

immuno-blotting with either anti-phosphotyrosine mAb, anti-c-kit serum, or

anti-PI3Kserum.

Immunecomplexkinase assay. Thecelllysateswereprepared from rhSCF-unstimulated cells using NP-40 lysis buffer.Theproductof c-kit

was immunoprecipitatedwiththeYB5.B8 mAb and protein

A-Sepha-rosebeads, andtheimmunecomplexes werewashed oncewithPBS,

twice with 0.5 MLiCl and50 mM Tris-HCl, pH 7.4,and once with

kinase buffer (10mMMnCl2,20 mMTris-HCl, pH 7.4)at4°C.Then, theimmune complexeswereincubated in kinase buffercontaining 10

;tCi/ml

[y-32P]ATP(DuPont/NENResearchProducts, Boston, MA) for 20 minat25°C, washed,andseparatedby SDS-PAGE.Thegelwas

dried, and radioactive proteinsweredetectedby autoradiography. Oligonucleotide primers. PCR was performed by usingtwo comple-mentaryoligonucleotide primers for amplifying human SCF cDNA:

5'-ATGAAGAAGACACAACTTG-3'(184to203,sense)and

5'-AAG-GCATCAATGGATCTATT-3'(615to634, antisense).Forthe ampli-fication ofhuman c-kit cDNA, the following seven oligonucleotide primerswere used:primer 1, 5'-GATCCCATCGCAGCTACCGCG-3' (1 to21,sense); primer 2,5'-ATCTTCCCCATGATAAACAC-3'(964

5'-AGTGTGCTCAGAAAGACAGGATTGC-3' (3068 to3092,

anti-sense).The numbersinparenthesesrepresent the nucleotide numbers

onthecomplementarystrandsof each cDNA sequence.

cDNAsynthesisand PCRamplification. Total cellular RNAwas

isolated with aguanidine isocyanate method in combination with a

cesium chloride modification(40, 41). GenomicDNA wasprepared

from HMC-Icellsbythe standard method. Total RNAwasconverted into cDNAusingtheprimerspecificfor thesubsequentPCR

amplifica-tion. cDNAsynthesizedfrom 4Agof total RNAwasamplifiedina100

Mgl reaction mixture by aDNA thermal cycler (Perkin-ElmerCetus

Corp., Norwalk,CT)usingTaqDNApolymerasein 25cyclesof1-min

denaturationat

940C,

2-minannealingat

550C,

and3-minsynthesisat

720C.To minimizepossibleartifacts that could be caused by priming withPCR,the number of PCRcycleswaskepttoaminimum.

DetectionofhumanSCF mRNAtranscriptbymeansofPCRafter

reversetranscription(RT-PCR) analysis.Thesingle-strandcDNAwas

synthesizedfrom total mRNAof HMC-I andNTI-4cellswith both the

specificand random primers, andsubjected toPCR toamplify the 45l-bpDNAfragmentbetweennucleotide 184 and 634 ofhuman SCF

cDNA.PCRproductswereexaminedon1% agarosegel,transferredto

nitrocellulose filter(Schleicher&Schuell, Dassel, Germany),and then

hybridizedwith random 32P-labeledprobe.

Sequencingofc-kit cDNA.Sequencingof the wholecoding region

ofc-kitcDNAwas_performedinHMC-I cellsby usingfoursetsof

primers: primers 1-4, 2-6,3-6,and 5-7asdescribed above. The PCR

products amplified by these primers were gel purified, sequentially

treatedwith T4polynucleotide kinase and the Klenow fragmentof

DNApolymerase I,andsubclonedinto the EcoRV site ofBluescript KS(-) (Stratagene Corp.,LaJolla, CA).DNAsequences of four sub-clonesobtained from each PCRproductweredetermined and

com-pared with normal sequence of human c-kit cDNA(2). TheDNA

sequencingreactionwasperformedbythemethodofSangeretal.(42)

using

[32P]a-deoxycytidine

triphosphate (DuPont/NEN Research

Products).Inordertofurther confirm the mutations of c-kit mRNA, thesequencingin theregionsofc-kitmutationswasperformed directly

usingDNAfragmentsamplifiedbyindependentRT-PCR.

Site-directed mutagenesis andtransfection.Todirectlyexamine the causal roleof thec-kit mutations in activation of c-kittyrosinekinase

activity,site-directedmutagenesiswasperformed usingmurinec-kit cDNAbecauseof thefollowingreasons._First,amino acid sequences in theregionsof thetwoc-kit mutationsarecompletelyconserved

be-tweenmouseand human.Second,human SCFis knowntohavelittle effectoncellsexpressingmurine c-kit receptor(c-kitR) (6,7), whereas murine SCFcanexertits actionsthroughinteraction with both murine and humanc-kitRs (6, 7, 27, 43).Wetherefore usedacombination of murinec-kit cDNA andahuman293T celllineas atransfectant.

The geneencodingmurinewild-type c-kitRwasclonedinto

Blue-script IKS(-) andthe orientation of theinsertwasdetermined by

sequencing. Single-stranded uracil-containing DNA was prepared fromEscherichia colistrain CJ236(dut-ung-)asdescribed by Kunkel

(44).To generate genescontainingG-559 and V-814 mutations, two

oligonucleotidesof 5

'-CAATGGAAGGTTGGCGAGGAGATAAAT-3' and 5'-GGGCTAGCCAGAGTCATCAGGAATGAT-3'were

syn-thesized,annealedtothetemplateandextended. The double-stranded mismatchcontainingmutantplasmidsweretransformed into ung+ E. colistrain,and the clonescontainingeachmutation were then screened

bysequencing. TheDraI-BssHIIfragmentwith G-599 mutation and theNheI-MroI fragmentwith V-814mutationwereisolated and in-serted into thewild type c-kit clonetoconstruct mutantsencoding

G-559,V-8

14,

orbothof them. The fullcodingsequences of the wild-type andmutatedc-kitcDNA werethen released bySmaIandHindIII

restrictiondigestion, isolated,andinserted into the expression vector

pSV2. Theinsertswere resequencedtoconfirm the orientation and mutations within theexpressionvectors.

The expression vectors containing wild-type and mutated c-kit

cDNA were transfected into 293T cells by the calcium phosphate method as described previously (45). The cells were cultured in DMEMcontaining 10% FCS and fed every 3 d with fresh medium. To

determine the expression of c-kit product in the transfected 293Tcells, the transfectantswere incubated with 100MCi/ml [35S]methionine (DuPont/NEN ResearchProducts) for 6h, andlysedwith NP-40lysis

buffer. The labeled c-kit product was then immunoprecipitated with a ratanti-mouse c-kit mAb (ACK-2) and analyzed by SDS-PAGE and autoradiography. For the analysis of protein tyrosine phosphorylation in thetransfectants, cell lysates were subjected to SDS-PAGE and

im-munoblottingwithananti-phosphotyrosine mAb. To examine the

ac-tivationandtyrosine phosphorylationofc-kitproduct,c-kitproductof each transfectant was immunoprecipitated with rabbit antiserum against murine c-kitproteinfrom thelysates, and then subjected to in vitro immune complex kinase assay and immunoblotting with an

anti-phosphotyrosinemAbasdescribed above.

Results

Constitutivetyrosine phosphorylation ofc-kit product in HMC-Icells. The expression of c-kit productwasfirst examined with

ananti-c-kit mAb,YB5.B8, which recognizes the extracellular domain of human c-kitproduct(35). Flow cytometry analysis showed that c-kitproduct wasexpressedon HMC-1, M07E, and HEL cells at an almost equal level (Fig. 1 A). Despite the high level of c-kit expression onHMC- 1 cells, proliferation of HMC-I cells wasnotaugmented by thetreatmentwithrhSCF

(datanotshown), whereas rhSCF was reportedtoinduce prolif-eration of M07E andHELcells(27,31).

Toexaminethe impaired effect of rhSCF on proliferation of HMC- 1 cells, we investigated the effectofrhSCFontyrosine kinase activity in HMC-1, M07E, and HEL cells. The cells were removedfrom growthfactors for16 h and thenstimulated

with 100 ng/ml of rhSCF for 15 minat 37°C. Changes in

tyrosine phosphorylation were detected by immunoblotting

withamAbspecific for phosphotyrosine.After stimulation of M07E and HEL cells with rhSCF, increased phosphotyrosine was observedinproteins particularlyatmolecularmassof145 kD(Fig. 1 B). In HMC- 1 cells, by contrast, tyrosine

phosphor-ylation of 145-kD protein was observed even in the absence of exogenousrhSCF, and rhSCF treatment resulted in only a modest increase in tyrosine phosphorylation of the protein

(Fig. 1 B).

Inorder todetermine ifthe 145 protein wasc-kitproduct,

c-kitproductwasimmunoprecipitatedand analyzed by

immu-noblotting with either anti-phosphotyrosinemAb or anti-c-kit serum.Asshown inFig. 1 C, tyrosine phosphorylationof c-kit

productwasdetectedinHMC-Icellsregardless ofrhSCF

stimu-lation, while itwasconsiderably increasedbystimulationwith

rhSCFinM07E andHELcells.

Activation

of

c-kitproduct in HMC-J cells. Ligand binding

toc-kit product is knowntoactivate the c-kitreceptortyrosine kinase, thereby leading to c-kit-tyrosyl autophosphorylation

andtheassociation of c-kit with itssubstratessuch as

phospha-tidylinositol 3-kinase (PI3-kinase)(4, 46-48). To determine whether or notthe constitutive tyrosine phosphorylation of c-kitproductin HMC-l cells wasaccompanied by activation of

c-kitreceptortyrosine kinase,weexamined the association of

c-kit product with P13K andthe autokinase activityofc-kit products inthe cells.

HMC-1, M07E,and HELcells were factor starved and

ex-posed

tomedium aloneorrhSCF(100ng/ml)for 15min at

37°C, and then lysed with lysis buffer containing 1% digitonin. Beforeandafter stimulationwith rhSCF, c-kit products were

immunoprecipitated withaYB5.B8 mAb from the cell lysates

anddivided intotwoaliquots, oneof which was subjected to

A

a)

E HMC-1

M07E

HEL

:U

Relative

log

fluorescence

HMC-1

M07E

HEL

I

m

m

SCF

- + - + -

+

cc-P13K

Z-0

*--85kD

B

HMC-1 MO7E HEL

SCF - + -+

MW(kD)

208-

100-

71-C

HMC-1 M07E HEL

SCF - + - + - +

cc-PTyr

S"~ t~ 5S14k--4 _-*-145kD

cx-c-kit a l *-145kD

Figure1.Expression of c-kit productonHMC-I, M07E,and HEL

cells(A)andeffectsofrhSCFontyrosine phosphorylation ofcellular

proteins (B) and c-kit product (C) inthe celllines.Flow cytometric

analysisofthe surfacebinding ofamonoclonal anti-c-kit antibody

(YB5.B8)toHMC-1, M07E,and HEL cells (A). Cellswereincubated

in eithernegative control antibody(---)orYB5.B8(-), washed,

incubated withfluorescein-conjugatedgoat(Fab')2anti-mouseIgG

antibody,andanalyzedonaFACScanO.Foridentificationof

phos-photyrosine-containing proteins, celllysateswereobtained before and

aftertreatmentwith rhSCF (100 ng/ml) for15 minat370C and

subjectedtoimmunoblot with anti-phosphotyrosinemAb (B).To

examinechangesintyrosine phosphorylation of c-kitproduct,c-kit

productwasimmunoprecipitatedwith YB5.B8fromcelllysates

be-fore and 15min after rhSCF stimulation,andimmunoblotwas

performed with eitheranti-phosphotyrosinemAboranti-c-kit serum(C).

immunoblottingwith anti-PI3Kserumand another with

anti-c-kitserum.Basedona-c-kitimmunoblotdataasshown in the

lower portion of Fig. 2, approximately equivalentamountsof c-kit productswereimmunoprecipitatedbefore and after

stimu-lation with rhSCF in each cell line.Bycontrast,a-PI3K immu-noblotasshown in theupperportion of Fig. 2revealed that the

85-kD subunit of P13Kwasco-immunoprecipitatedwith

anti-c-kit antibody only afterstimulation with rhSCF in twocell

lines, M07E and HEL, whereas itwasco-immunoprecipitated

in HMC-1 cellsevenbeforestimulationwith rhSCF.

Wenextexamined the autokinase activity of c-kit products

in immune complex kinase assay.Productsof c-kitwere

im-munoprecipitatedfrom NP-40 lysates of HMC-1, M07E,and

HEL cellsbeforetreatmentwith rhSCF,and assayedfor

auto-kinase activity. Tyrosine auto-kinase activity of c-kitproductswas

ct-c-kit

1

*1

45kD

0MR1 _Om* ON _-am

Figure2. Association of c-kitproductwith P13K inHMC-1,M07E,

and HEL cells. Cellsweretreated with rhSCF(100 ng/ml)for 15 minat_370Candlysedwith 1%digitonin.c-kitproductand c-kit-as-sociated moleculeswereimmunoprecipitatedwith YB5.B8 and

pro-teinA-Sepharosebeads from the celllysatesbefore and after stimula-tion with rhSCF. Theimmunoprecipitatesweredivided into two ali-quotes,separatedby SDS-PAGE,and subjected to immunoblot

analysesof P13K(upperpanel) and c-kit product (lowerpanel).

readily

detectablein

HMC-l

cellsin the absence of SCF

stimu-lation,whereas itwas

only

faint inM07EorHELcells(Fig.

3).

In this experiment, almost equal amounts of c-kit

products

wereimmunoprecipitatedinthese_{samples (data}not

shown).

Absence ofSCF mRNA transcripts in HMC-I cells. One of thepossible explanationsfor the results shown inFigs. 1-3was

thatHMC- cells

might

produceSCF. Theexpressionof SCF mRNA transcriptswas examinedby RT-PCR Southernblot

analysisinHMC-1 cells and also in NTI-4 cells thatwereused asapositivecontrol.As shown inFig. 4,SCF mRNA

transcript

wasnotobservedinHMC-1 cells,whilea45l-bp fragmentof

SCFmRNA was

easily

detected in NTI-4 cells. Further, the culture supernatant ofHMC-l did not haveanyactivities in

stimulating proliferation

or

inducing

tyrosine

phosphorylation

of c-kit product in M07E cells(datanot shown). Thus, the constitutive tyrosine phosphorylation andactivation of c-kit

productinHMC-1cellswerenotattributedtoautocrine mech-anism.

N> /Figure3. Immune

com-ti/ A\> plexkinaseassayfor

c-Ov/

of/wt>

kit tyrosine kinase

ac-tivityinHMC-l,M07E,

and HEL cells. c-kit product was

immuno-MW (kD)

precipitated

with

YB5.B8 from NP-40

ly-208- sates before stimulation withrhSCF, andthe c-kit -* immunoprecipitates

U $^3 wereincubated in

ki-nasebuffer containing

1

00

10

ACi/ml

[y-32P]ATP

for 20minat250C,

washed, and separated 71- by SDS-PAGE.

Incor-poration of 32P into c-kit productwasvisualized by autoradiography. In thisstudy, theimmunoprecipitatesfrom HMC- 1,M07E,and HEL cells containedapproximately equivalentlevelsof c-kit products.

Figure 4. RT-PCR anal-ysis of human SCF

mRNAinHMC-l and NTI-4 cells.

Single-X\ strand cDNAswere

2j/\/>/synthesizedfrom total

\>28> +N/ mRNA ofHMC-l and

C _{NTI-4cells, and the}

45l-bpfragment of

hu-manSCF cDNA (nu-cleotides 184 through 634)wasamplified by PCR. The PCR

prod-451bp30 ucts weresubjected to

451 bp

*

_

l%

agarose gel,

trans-(184- 634) 7 ferredtonitrocellulose

filter, hybridized with random32P-labeled

probe, and then de-tected by autoradiog-raphy.

Mutations inc-kit cDNA. An alternate mechanism which

might explain the constitutive tyrosine phosphorylation and

activation of c-kitproducts in HMC-1 cells would be due to

mutations inthec-kitproto-oncogene. Southern blot analyses

ofgenomic DNA from HMC-1 using EcoRI, HindIII, and BamHI showedneitherrearrangement nor amplification in

c-kitgene as comparedwithnormal human placenta (data not

shown). However, sequencing of c-kit cDNA demonstrated thatc-kitcDNAofHMC-1 cells carries heterozygous transi-tions in codon 560

(GTT

--

GGT,

nucleotides 1699

through

1701 intheregion between the membrane and the

ATP-bind-ing domains)

and codon816

(GAC

GTC,

nucleotides2467

through 2469 inthephosphotransferasedomain), which result

intheamino acid substitutions ofGly-560 for Val and Val-816

for Asp,respectively.Thesetwopointmutationswereobserved in cDNA clonesgeneratedfrom threeindependentPCR

prod-ucts, and also confirmedbydirectsequencingofPCR

products

amplified independently (Fig. 5). Further, sequencing of cDNAfragment (nucleotides964through 2594) coveringthe regions of both mutations was performed in four subclones. This analysis showed that two subclones had bothmutations,

but theremainingtwosubcloneshad neither mutation. These results indicate that c-kit genes ofHMC-1 cellsarecomposedof

anormal wild-typeallele anda mutant allelewith two

point

mutations.

Inaddition, the deletions of 12-nucleotide sequence

(nu-cleotides 1550through 1561),the extracellular4-amino acid sequence(Gly-Asn-Asn-Lys),and3-nucleotidesequence

(nu-cleotides 2163 through 2165) were observed in both normal and mutant alleles of HMC- 1. The former deletion was

previ-ously reported to be due to alternative splicing of mRNA(49);

thelatter was also consideredasthe result ofalternative

splic-ing of mRNA between exon 14 and 15, in light of data on the humanc-kitgenome(50).

Mutation(s) leads to activation of c-kit receptor tyrosine kinase. Amino acidsequencesinthe regions ofthe twoc-kit

mutations identified in HMC-l cells arecompletely conserved in all ofmice, rats, and humans. In order to determine the direct role of the c-kit mutations in ligand-independent activa-tion of c-kit products, the expression vectors containing

nor-mal ormutatedmurine c-kitgenes weretransfected into cellsof a humanembryonic kidney cell line, 293T. This combination could minimize the effect of SCF, which might be producedby

the transfected cells, on c-kit product expressed on 293T cells (see Methods).

Weused site-directedmutagenesisto construct mutant

mu-rine c-kitR containing substitutions of Gly forVal at codon

559and/orValforAsp at codon814,whichcorrespondedto Gly-560 and/or Val-816 of mutant human c-kitR,

respec-5'

554

-EG

Val[

T

A

Ginr

A

LG

T \C A T G

LGu

AA

LysE

AG

G[

Val[ETA

/T -_30 )y-

l/

Val

_[EQT

₃₁₁

G

Glu

I

G

-A_ _

lie

T

IPI

[A|

56-Asn[A

5'

A

_]Asp

_{- 810}

C

A

T

G

T

]Phe

G]Gly

T

T

]Leu

_- A

C'

c Ala

A

;G]Arg

A

G--

Aij

Asp/Val

-816 c

A-T

Ile

c

A-A]

Lys

G .

A-A Asn

T

G-A]

Asp

T

c

_Ser

T

A

]

Asn

-

822

Figure 5. Nucleotidesequenceincludingthe mu-tationsites of c-kit cDNAobtainedfromHMC-I

cells. Shownareresults from direct sequencing of c-kitDNAfiagmentsfromamino acid residues 554 to566 andthose 810 to 822. HMC-1 cells

werefound tocarryheterologousmutationsof

c-kitgeneat codons560 and816. These two muta-tionswereobserved incDNA clones from three independent PCR products, and sequencing of

cDNA fragments (nucleotides 964through 2594) infourclonesrevealed that the c-kitgeneof HMC-l iscomposed ofanormal andamutant allele. Thearrowsindicate the mutation sites, and thealterednucleotides and amino acidsare

un-derlined.

A

MW(kDa) K K K

9,

170-

116-

76-x9 0,

(5

*<--l

45k

Da

-<-l 25kDa

aFF

..

B

AW(kDa)

₍

216-

105-- 145kDa -<-l25kDa

Figure 6.Expressionofc-kitproduct(A)andidentification of

phosphotyrosine-containing proteins (B)

in 293T cells transfected withwild-type

and mutated murine c-kit genes. The cellsweretransfected withwild-typeormutatedmurine c-kit genes

encoding

wild-type

c-kitR

(wild

type),

c-kitR(Gly-559;G-559),c-kitR(Val-814;V-814),orc-kitR(Gly-559,Val-814; G-559, V814). Also,293T cells that didnotreceive transfection wereused as anegative control (notransfection).(A) The indicated cellswerelabeledfor 6hwith 100IlCi/ml

[35S]methionine

andlysed.

c-kit productwasthenimmunoprecipitatedandanalyzedbySDS-PAGE andautoradiography. (B) Lysatesof the indicated cellswerenormalized forproteincontent, andweresubjectedtoSDS-PAGE andimmunoblottingwithan

anti-phosphotyrosine

mAb.The mobilities of the mature (145-kD) and immature(125-kD)forms of c-kitproductsareindicatedatright.

tively.Whenwild-typeormutatedc-kitgenes were transfected

in 293Tcells, thec-kitgene products weresynthesizedin the

cellsas 145- and 125-kDproteins(Fig. 6 A), whichwere the mature and immature forms of c-kit products, respectively

(51). Although each transfectant expressed approximately equal amounts ofc-kit products, a major c-kit product that

incorporated

[35S]methionine

wastheimmature form inthe cellsexpressingc-kitR(Gly-559, Val-814)orc-kitR(Val-814),

whereasitwasthe mature formin the cellsexpressing c-kitR (Gly-559)orwild-type c-kitR (Fig. 6 A).

Beforeandafter transfection ofc-kitgenes,changes in

tyro-sine

phosphorylation

ofproteins

weredetected by immunoblot-tingwith

anti-phosphotyrosine

mAb

(Fig.

6B).Ascompared with nontransfected 293Tcells,littleor noincreaseintyrosine phosphorylationwasobservedin the cells expressing wild-type

c-kitR. However, a number ofproteins,

including

125- and

145-kDproteins, wereheavily phosphorylatedontyrosinein the cellsexpressingc-kitR(Val-814)orc-kitR (Gly-559,

Val-814). Also, a modest degree of tyrosine phosphorylation of 125- and 145-kD proteins was notedin the cells expressing

c-kitR (Gly-559).

A

B

-<*-145kDa

_i _1 <(-1 25kDa

Inordertofurther examine thetyrosine

phosphorylation

and activation of c-kitproductsin the transfected cells,c-kit productswereimmunoprecipitatedand assayed by immuno-blottingwithananti-phosphotyrosinemAb and also for

auto-kinase activity. In the absence of SCF, both c-kitR(Val-814)

andc-kitR(Gly-559, Val-814)were foundtobe abundantly

phosphorylated on tyrosine, and c-kitR (Val-814) was also phosphorylated on tyrosine, albeit to a much lesser degree, whereaswild-typec-kitRwasnot(Fig. 7 A). Further, in accord with the findings onthe immunoblotting analysis, immune

complexkinaseassayshowedastrikingautokinase activity in both c-kitR(Val-814)and c-kitR(Gly-559, Val-814),aweak activity in c-kitR(Gly-559),and littleor noactivity in

wild-typec-kitRin the absence of SCF (Fig. 7B).

Discussion

Many receptor tyrosine kinases were first identified as

retro-viral oncogenes such as v-erbB,v-fms, andv-kit thatencode

oncogenicforms of epidermal growth factor R (EGFR), CSF-1

R,

andc-kitR,respectively(1, 46, 52). Comparative

nucleo-:1r

l45kDa 125kDa

Figure 7. Tyrosine phosphorylation (A) and acti-vation (B) of wild-type and mutated murine c-kitR expressed in 293T cells in the absence of ex-ogenous SCF. (A) c-kit product was immunopre-cipitated from the lysates of the indicated cells, andsubjectedtoSDS-PAGE and immunoblotting with an anti-phosphotyrosine mAb. (B) Immune complex kinase assay for c-kittyrosinekinase

ac-tivity. c-kit productwasimmunoprecipitatedfrom lysates from theindicated cells, and the immuno-precipitateswereincubated in kinase buffer

con-taining 10

gCi/ml

[y-32P]ATPfor 20 min at 25°C, washed, and separated by SDS-PAGE.

In-corporationof 32P intoc-kit productwas visual-ized byautoradiography. Themobilities of the

mature(145-kD) and immature (125-kD) forms of c-kitproductsareindicatedatright.

Activating MutationofProto-oncogene c-kit 1741

tide sequencing analysesrevealed avarietyofstructural

differ-ences between the oncogenes andtheir proto-oncogenes:

v-erbB

andv-kit proteins lacktheextracellular binding domain of the receptorsthatnegatively regulatesthereceptor

tyrosine

kinase activity; v-fms differs from c-fms by nine scattered

amino acid substitutionsand byaCOOH-terminal structure

(3, 46, 53, 54). These structural alterations in the oncogenes activate constitutively the tyrosine kinase function,

thereby

leadingtocelltransformation and

tumorigenicity

(46).In

ad-dition,

it hasbeen reported that even point mutations in a

proto-oncogene can lead to activation of the corresponding tyrosine kinase receptor, as has been shown by

"activating"

mutations of c-fms that induce morphologic transformation,

anchorage-independent

growth, and

tumorgenicity

in mouse

NIH 3T3 cells (53, 54). In the case of c-kit, however, all of

mutations so far describedin mice, rats, and humansare

re-ported to affect either the structure of c-kitorc-kit expression, resulting in a decreaseinc-kit tyrosine kinase activity (4).

In this study, we have shown that c-kit product is constitu-tively phosphorylated on tyrosine and activated in HMC-1 cells even in the absence of SCF-stimulation. Further, c-kit productwasfoundto beassociated with PI3K without the exog-enous SCF. The ligand-induced association ofc-kit product with P13K is reported to depend on

autophosphorylation

of c-kit product on tyrosine residues where the associationoccurs

(48). Therefore, tyrosine residuesthat are

constitutively

phos-phorylated in HMC- 1 cells may be indistinguishable from those phosphorylated byligand stimulation. However, neither mRNA transcript nor biological activity of SCF were detect-ablefrom HMC-1cells. These results suggest that c-kit product in_{HMC-1 cells is constitutively activated in a} ligand-indepen-dent manner.

Sequencing analyses have revealed that c-kit genes of

HMC-l cellsarecomposedof anormal, wild-type alleleanda mutant allele with two point mutations resulting in amino acid substitutions of_{Gly-560 for Val and Val-816 for Asp in the}

intracellular domain. In order to analyze the effect ofthese previously unidentified substitutionsonc-kit kinase function, mutated murine c-kit

genes encoding murine c-kitR

(Gly-559 and/or Val-814), which were constructedbysite-directed mu-tagenesistocorrespondtohuman

c-kitR

(Gly-560 and/or Val-816), were transfectedin ahuman embryonic kidney cell line, 293T cells.In thetransfected cells, wild-type

c-kitR

was neither phosphorylated ontyrosine invivo nor activated inimmune complex kinase reaction in theabsence of SCF. By contrast, both c-kitR (Gly-559, Val-814) and c-kitR (Val-814) were strikingly phosphorylated ontyrosine and activated in the ab-sence ofSCF. Further, c-kitR (Gly-559)was also phosphory-lated on tyrosine and activated, albeit to a much lesser degree. Given the fact that human SCF is hardly active on murine c-kitR(6, 7)evenif itmightbeproduced from 293T cells, these results indicatethatconversionofaspartic acid-814tovalinein

murine c-kitRis anactivating mutationandconsiderably more efficient than that ofvaline-559 to glycine in inducing c-kit kinase activity.It istherefore suggestedthat amutationof

hu-man c-kit geneatcodon816(Asp -* Val) playsamajor rolein

the constitutive activation of c-kit product in

HMC-1

cells, whileamutationatcodon 560 (Val -- _{Gly) hasaminor role.}

The activating mutation of

c-ffms

geneisreportedtoaffect intracellular transport ofhuman

CSF-1R

(54). Whena wild-type human c-fms gene was transfectedintomurine NIH/3T3

cells,

theimmature formofhuman

CSF-lR

(gpl

30

_'f?)

was

rapidly

convertedtothematurecell surface formof

CSF-IR

(gpl50 'fm). By

contrast,

whenamutatedhuman

c-fms

gene

bearing

an

activating

mutation

(Leu-301

--

Ser)

_was

trans-fected,

theconversion fromtheimmaturetothematureform

ofCSF-IRwas

_{inefficiently}

_processed.

_{In this}

_study,

_wefound

that theimmature form

(

125

kD)

ofc-kitRwas

_{predominantly}

observedin cells

_expressing

c-kitR

(Gly-559, Val-814)

or

c-kitR

_{(Val-8 14),}

whereasthe matureform

(145

kD)

of

c-kitR

was

_predominant

in cells

expressing

c-kitR

(Gly-559)

or

wild-type

c-kitR. Theseresults

_provide

additional

support

for the ideathat Val-814 in murine

c-kit, corresponding

toVal-816 in

humans,

isan

_activating

mutation.

Ithas been

reported

that

only

alowbutdetectable levelof

constitutive activationisnoticeableinthe shorter formof

mu-rine c-kit

_product,

which lackstheextracellular4aminoacid sequence

(Gly-Asn-Asn-Lys)

duetothealternative

splicing

of

mRNA

_(49).

This shorter formofc-kitisfoundinthe halfof normal murine or rat c-kit cDNA clones

(40, 49).

Also in

HMC-l

cells,

wefoundthat thehalfof c-kitcDNA clones lacks

thesamefour-aminoacid sequence inbothnormal and

mu-tated c-kitgenes.In ordertocompare c-kitR

(Gly-559 and/or

Val-8

₁₄₎

with theshorter form of

wild-type

c-kitRand to

de-termine thedirectrole ofthec-kit mutationsin theconstitutive

activation,

weused theshorter formof

wild-type

murine c-kit

cDNA as a

_template

_{for the site-directed mutagenesis} in this

study.

In thetransfected 293T

cells, tyrosine

phosphorylation

and activation were

_barely

detectable in the shorter form of

c-kitR

whichwasusedas anormal

control,

whereas

they

were

readily

detectablein

c-kitR

(Gly-559

and/or Val-814).

These

resultsindicatethat c-kitmutations identifiedin this

study

are

considerably

moreactive than the shorter form of

c-kit,

and that theshorter formis

unlikely

tocontributetothe

constitu-tiveactivationofc-kit

product

inHMC-1 cells.

Although

themutations described herearefoundtoleadto

theconstitutive activationof c-kit

tyrosine

kinase activity,the role of the c-kit

activating

mutations in leukemogenesis

re-mainstobe elucidated. Ina

previous study,

we showed that

c-kit

_product

was

phosphorylated

on

tyrosine

in

freshly

pre-pared

leukemiccells ofsomeAMLcases evenin theabsenceof

SCF stimulation

(28). However,

occurrencesof

tyrosine

phos-phorylation

of c-kit

product

observedin theAMLcaseswere not so

striking

asthat inHMC-1 cells

(28). Therefore,

an

acti-vating

mutationatresidue 816

(Asp

-- _{Val) maynotoccur}

frequently

in denovo

AML,

while it is still possible that certain mutations of

c-kitR

otherthanc-kitR (Val-8 16) may

contrib-ute tothemodest

degree

of c-kit activation in AML cells. It is

interesting

to notethat HMC-l is a human mast-cell leukemia

cellline

(37),

andthatamouse mastocytoma cell line, P-8 15, hasalsobeen reportedto exhibit constitutive activation of c-kitRin the_{absence of detectable autocrine production of SCF}

(48), although

the_{sequence of the c-kit alleles expressed by}

P-815cells isnot_{characterized at this time. Further, a}

_c-kitR/

SCF system isknownto

play

acrucial role in the growthand

differentiation ofmastcells (4, 16). It has been recently

re-ported

that_{apparently selective growth of human mast cells is}

observedin

long-term

culture of human cord-blood or

bone-marrowcells

exceptionally

_{with either murine or human SCF} butnotwithother_{hematopoietic growth factors (43, 55, 56).}

Also,

von Ruden et al._{have recently reported that transfer of}

mice results inthe development of lethal systemic mast cell

leukemia, and suggested thatv-erbBmay use partsofa signal

transduction pathwaynormally coupled toc-kitR(57). These

findings raisethe possibility that activating mutations of c-kit

describedhere may beinvolvedin neoplastic transformation of mast cells, althoughit isas yet unclearwhetherthesemutations by themselvescan leadtofactor-independentgrowthand

tu-morgenicity. It would beofconsiderable interest to examine whether thetransfection of nontumorigenic or

factor-depen-dent cell lineswith the mutantforms ofc-kitgenesresults inthe

production ofatumorigenicor factor-independent phenotype. In addition to hematological malignancies, expression of

c-kit

proto-oncogene has been found in varioustumorssuch as lungcancer and germ cell tumors (58, 59). Further, the amino

acidsequenceinaregionofthe kinase domain containing

Asp-816ofhumanc-kit product is widelyconservedinmany other RTKs such as receptorsfor PDGF, CSF- 1, andinsulin(60).

Therefore, it is possiblethat the uniqueactivating mutationin codon 816 of human c-kitR, or a similar mutation in other RTKs, mayparticipate inthedevelopmentofavarietyof

ma-lignancies.Furtherstudiesonc-kitmutants will be necessary to

understand themechanisms by which growth factorreceptors influenceoncogenesisand normal growth control.

Acknowledgments

Wewould like to thank Dr. K. M. Zsebofor generously providing rhSCF, Dr. A. Ullrich for rabbit anti-human c-kit serum, Dr. P. Besmerfor rabbit anti-murine c-kit serum,Dr.S-I.Nishikawa for rat anti-murine c-kit mAb (ACK-2) and full length of murine c-kit cDNA

clone,and Dr. D.Baltimore for 293T cell line.

This workwassupported in part by grants from theMinistryof Education, Science and Culture ofJapan, the Research Foundation for Cancer and Cardiovascular Diseases(Osaka),andtheInamori Foun-dation.

References

1.Besmer,P., J. E. Murphy, P. C.George,F.Qui,P.J.Bergold,L.Lederman, H. W.Snyder Jr., D.Brodeur,E. E.Zukerman, andW. D.Hardy. 1986.A new acutetransforming feline retrovirusandrelationshipof its oncogene v-kit with theproteinkinase genefamily.Nature(Lond.).320:415-421.

2.Yarden, Y., W. J. Kuang, T. Yang-Feng, L. Coussens, S.Munemitsu,T. J. Dull,E.Chen, J.Schlessinger,U.Francke, andA.Ullrich. 1987. Human proto-oncogenec-kit:a newcellsurfacereceptortyrosinekinase foranunidentified ligand.EMBO(Eur.Mol.Biot.Organ.)J. 6:3341-3351.

3.Qiu,F.H., P. Ray, K. Brown,P. E.Barker, S.Jhanwar,F. H.Ruddle,andP.

Besmer. 1988.Primarystructureofc-kit:relationshipwith theCSF-1/PDGF

receptor kinasefamily-oncogenicactivation ofv-kit involves deletion ofextracel-lulardomain and C terminus. EMBO(Eur. Mol.Biot. Organ.)J.7:1003-1011.

4.Reith,A.D.,andA.Bernstein. 1991. Molecularbiologyofthe WandSteel loci.InGenomeAnalysis.Volume3. Genes andPhenotypes.ColdSpringHarbor LaboratoryPress, ColdSpring Harbor,NY.105-133.

5.Zsebo, K. M., J. Wypych, I. K.McNiece,H. S. Lu,K. A.Smith,S. B. Karkare,R.K.Sachdev,V. N.Yuschenkoff,N.C.Birkett,L. R.Williams,etal. 1990.Identification,purification,andbiologicalcharacterization of

hematopoi-eticstemcellfactor from buffaloratliver-conditionedmedium.Cell. 63:195-201.

6.Martin,F.H.,S.V.Suggs,K.E.Langley,H.S.Lu,J.Ting,K.H.Okino,

C.F.Morris,I.K.McNiece,F. W.Jacobson,E. A.Mendiaz,etal.1990.Primary structureandfunctionalexpressionofratand human stemcell factorDNAs.

Cell.63:203-211.

7.Zsebo,K.M.,D. A.Williams,E. N.Geissler,V. C.Broudy,F.H.Martin, H. L.Atkins,R. Y.Hsu,N.C.Birkett,K. H.Okino,D.C.Murdock,et al.1990. Stem cellfactor is encodedat theSIlocus of themouseandistheligandforthe

c-kittyrosinekinase receptor. Cell. 63:213-224.

8.Williams,D.E.,J.Eisenman,A.Baird,C.Rauch,V.K.Ness,C. J.March, L.S.Park, U.Martin,D. Y.Mochizuki,H.S.Boswell,etal.1990.Identification

ofaligand for thec-kit proto-oncogene.Cell. 63:167-174.

9.Copeland,N.G.,D. J.Gilbert,B.C.Cho,P.J.Donovan,N.A.Jenkins,D.

Cosman,D.Anderson,S. D._Lyman,and D.E.Williams. 1990. Mast cellgrowth

factor mapsnearthe steel locuson mousechromosome 10 andis deleted ina

number of steel alleles.Cell.63:175-183.

10.Anderson,D.M.,S.D.Lyman,A.Baird,J.M._Wignall,J.Eisenman,C.

Rauch,C. J.March,H.S.Boswell,S.D.Gimpel,D.Cosman,andD.E.Williams. 1990. Molecularcloningofmastcellgrowth factor,a_{hematopoietin}that is active in both membrane bound and solubleforms.Cell.63:235-243.

11.Nocka, K.,J.Buck,E.Levi,and P.Besmer. 1990. Candidateligandfor the c-kit transmembrane kinasereceptor.KL,afibroblast derivedgrowthfactor stim-ulatesmastcells anderythroid progenitors.EMBO(Eur.Mol.Biol.Organ) J. 9:3287-3294.

12.Huang, E.,K.Nocka,D. R.Beier,T.Y.Chu,J.Buck,H. W.Lahm,D. Wellner,P.Leder,and P. Besmer. 1990. Thehematopoietic growthfactorKLis encodedbytheSIlocusand is theligandofthec-kit receptor,the geneproductof

theWlocus.Cell.63:225-233.

13.Chabot, B.,D.A._Stephenson,V. M.Chapman,P.Besmer,andA.

Bern-stein.1988. Theproto-oncogenec-kitencodingatransmembranetyrosinekinase receptor mapstothemouseWlocus. Nature(Lond.).335:88-89.

14.Geissler,E.N.,M.A.Ryan,andD.E. Housman. 1988. The dominant-whitespotting (W)locus of themouseencodes the c-kitproto-oncogene. Cell. 55:185-192.

15.Russell,E. S. 1979.Hereditaryanemia ofthemouse: areview for

geneti-cists. Adv.Gen. 20:357-459.

16.Kitamura,Y. 1989.Heterogeneityofmastcells andphenotypic change

betweensubpopulations.Annu. Rev.Immunol.7:59-76.

17._Galli,S.J.,andY.Kitamura. 1987.Geneticallymast-cell-deficientW/W'

andS]IS]I_{mice.Am.J.Pathol. 127:191-198.}

18.Denberg,J.A.1992.Basophilandmastcelllineagesin vitro and in vivo.

Blood. 79:846-860.

19.Ogawa,M.,Y.Matsuzaki,S.Nishikawa,S-I.Hayashi,T.Kunisada,T. Sudo,T.Kina,H.Nakauchi,andS-I.Nishikawa. 1991.Expressionand function ofc-kitinhematopoieticprogenitorcells.J.Exp.Med.174:63-71.

20.Papayannopoulou,T.,M.Brice,V.C.Broudy,and K.M.Zsebo. 1991. Isolation ofc-kitreceptor-expressingcellsfrom bonemarrow,peripheralblood,

andfetalliver:functionalandcompositeantigenic profile.Blood. 78:1403-1412.

21.Ashman,L. K.,A.C.Cambareri, L. B.To,R.J.Levinsky,andC.A.

Juttner. 1991.Expressionof the YB5.B8antigen (c-kitproto-oncogeneproduct)

innormal humanbonemarrow.Blood. 78:30-37.

22.McNiece, I.K.,K.E. Langley,and K.M.Zsebo. 1991.Recombinant

humanstemcell factorsynergizeswithGM-CSF, G-CSF,IL-3 andEPOto

stimu-latehuman progenitorcellsof themyeloid erythroid lineages.Exp.Hematol.

(NY). 19:226-231.

23.Migliaccio,G.,A. R.Migliaccio,M.L.Druzin,P. J.V.Giardina,K.M. Zsebo,and J. W. Adamson. 1991.Effectsofrecombinanthumanstemcell factor

(SCF)onthegrowthof human_{progenitorcellsin}vitro. J.Cell.Physiol. 148:503-509.

24.Broxmeyer,H.E.,S. Cooper,L.Lu,G.Hangoc,D.Anderson,D. Cos-man, S. D.Lyman,andD.E. Williams. 1991. Effect of murinemastcellgrowth

factor(c-kitproto-oncogeneligand)on _colonyformationbyhumanmarrow

hematopoietic progenitorcells.Blood. 77:2142-2149.

25.Andre, C.,L.d'Auriol, C.Lacombe,S.Gisselbrecht,andF. Galibert. 1989.c-kitmRNAexpressioninhuman and murinehematopoieticcell lines.

Oncogene.4:1047-1049.

26.Wang,C.,J. E.Curtis,E. N.Geissler,E. A.McCulloch,and M.D.

Min-den. 1989. Theexpressionoftheproto-oncogene c-kitin the blast cellsofacute

myeloblasticleukemia. Leukemia(Baltimore).3:699-702.

27.Kuriu, A.,H.Ikeda,Y.Kanakura,J.D.Griffin,B.Druker,H._Yagura,H.

Kitayama,J.Ishikawa,T.Nishiura,Y._Kanayama,etal. 1991. Proliferation of

humanmyeloidleukemiacell line associated with thetyrosine-phosphorylation andactivationof theproto-oncogene c-kitproduct.Blood. 78:2834-2840.

28.Ikeda, H.,Y.Kanakura,T.Tamaki,A.Kuriu,H.Kitayama,J.Ishikawa,

Y._Kanayama,T.Yonezawa,S.Tarui,and J. D.Griffin. 1991.Expressionand functional role of theproto-oncogenec-kit inacutemyeloblasticleukemia cells.

Blood.78:2962-2968.

29.Buhring,H-J,A.Ullrich,K.Schaudt,C.A.Muller,andF. W.Busch. 1991. The product of the proto-oncogene c-kit

(P1451")

isahuman bone

marrowsurfaceantigenofhematopoieticprecursorcell which isexpressedon a subset ofacutenonlymphoblasticleukemia cells.Leukemia(Baltimore). 5:854-860.

30.Broudy,V.C.,F. 0.Smith,N. Lin,K.M.Zsebo,J. Egrie,andI. D.

Bernstein. 1992. Blasts frompatientswithacutemyelogenousleukemiaexpress

functionalreceptorsforstemcell factor. Blood. 80:60-67.

31.Pietsh, Y.,U.Kyas,U.Steffens,E.Yakisan,M. R.Hadam,Ludwig,K. M. Zsebo,andK.Welte. 1992. Effect ofhumanstemcell factor(c-kit ligand)on

proliferationofmyeloidleukemia cells:heterogeneityin responseandsynergy

with otherhematopoietic growthfactors. Blood.80:1199-1206.

expression in human myeloid- and mast cellprogenitors. J. Immunol. 147:4224-4228.

33. Kanakura, Y.,B.Druker,S. A.Cannistra,Y.Furukawa, Y. Torimoto, and J. D. Griffin. 1990. Signal transductionofthe human granulocyte-macro-phagecolony-stimulating factor and interleukin-3 receptors involves tyrosine phosphorylation of a common set ofcytoplasmic proteins. Blood. 76:706-715.

34. Kanakura, Y., B.Drucker, J. Decarlo, S.A.Cannistra,and J. D.Griffin.

1991. Phorbol 12-myristate acetate 13-acetateinhibits granulocyte-macrophage

colony-stimulating factor-dependentcellline. J.Biol. Chem. 266:490-496. 35. Lerner, N. B., K. H.Nocka, S. R. Cole, F. Qiu, A. Strife, L. K. Ashman, and P. Besmer. 1991.Monoclonalantibody YB5.B8identifiesthe humanc-kit protein product.Blood.77:1876-1883.

36.Majumder, S.,K.Brown, F-H.Qui,and P. Besmer. 1988.c-kit protein,a transmembranekinase: identification in tissues and characterization. Mol.Cell. Biol.8:4896-4903.

37.Butterfield, J. H., D.Weiler, G. Dewald, and G. J. Gleich. 1988. Establish-mentofan immature mast cell line from a patient with mast cell leukemia. Leuk. Res. 12:345-355.

38.Avanzi,G.C.,P.Lista,B.Giovinazzo,R.Miniero,G.Saglio,G. Benetton, R.Coda, G.Cattoretti, and L. Pegoraro. 1988. Selective growth response to IL-3

ofahumanleukemic cell line with megakaryoblastic features. Br. J.Haematol.

69:359-366.

39.Graham, F. L., J. Smith,W.C. Russell, and R. Nairn. 1977.

Characteris-ticsofahuman celllinetransformedby DNAfrom human adenovirus type 5. J. Gen. Virol.36:59-72.

40.Tsujimura,T., S.Hirota, S. Nomura, Y. Niwa, M. Yamazaki, T. Tono, E. Morii, H. M. Kim, K. Kondo, Y. Nishimura, and Y. Kitamura. 1991. Character-ization of Ws mutant allele of rats: a 12-base deletion in tyrosine kinase domain ofc-kit gene.Blood.78:1942-1946.

41.Tsujimura, T., U.Koshimizu, H. Katoh, Y. Isozaki, Y. Kanakura, T. Tono, S.Adachi, T. Kasugai, H. Tei, Y. Nishimune, et al. 1993. Mast cell number in theskin ofheterozygotesreflectsthe molecular natureofc-kitmutation.Blood.

81:2530-2538.

42. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain terminating inhibitors. Proc.Natl. Acad.Sci. USA.74:5463-5467.

43.Mitsui,H., T.Furitsu, A. M. Dvorak, A-M. A.Irani, L. B. Schwartz, N. Inagaki, M. Takei, K. Ishizaka, K. M. Zsebo, S. Gillis, and T. Ishizaka. 1993. Development of human mast cells from umbilical cord blood cells by recombi-nant human andmurine c-kitligand. Proc.Natl. Acad.Sci. USA. 90:735-739.

44.Kunkel,T.A.1985.Rapidandefficientsite-specific mutagenesis without

phenotypicselection. Proc.Natl. Acad.Sci. USA. 82:488-492.

45.Graham,F.L., and A. J. van der Eb. 1973. A new technique for the assay ofinfectivityof human adenovirus 5 DNA. Virology. 52:456-467.

46.Ullrich,A., and J.Schlessinger. 1990.Signaltransduction by receptors with tyrosine kinase activity.Cell. 61:203-212.

47.Koch, C. A., D. Anderson, M. F. Moran, C.Ellis, and T. Pawson. 1991. SH2 domains: elements that control interactions of cytoplasmic signaling pro-teins. Science(Wash. DC). 252:668-674.

48.Rottapel, R., M. Reedijk, D. E. Williams, S. D. Lyman, D. M. Anderson, T.Pawson, and A. Bernstein.1991. ThesteellWtransductionpathway: kit auto-phosphorylation and its association with a unique subset ofcytoplasmic signaling proteins is induced by the steel factor. Mol. Cell. Biol. 1 1:3043-3051.

49.Reith,A.D., C. Ellis, S. D. Lyman, D. M. Anderson, D. E. Williams, A. Bernstein, and T. Pawson. 1991. Signal transduction by normal isoforms and W mutantvariants of the Kit receptor tyrosine kinase. EMBO(Eur.Mol. Biol.

Organ)J.9:2451-2459.

50. Andre, C., E. Martin, F. Cornu, W-X. Hu, X-P. Wang, and F. Galibert. 1992.Genomicorganization ofthe humanc-kit gene: evidence of the receptor tyrosine kinase subclass III. Oncogene. 7:685-691.

51.Nocka, K., S.Majumder, B.Chabot,P.Ray, M. Cervone, A. Bernstein, and P. Besmer. 1989.Expressionof c-kit geneproductsin known cellular targets ofWmutations in normal and Wmutantmice-evidence foranimpairedc-kit kinase in mutant mice. Genes & Dev. 3:816-826.

52.Yarden, Y., and A.Ullrich. 1988. Growth factor receptor tyrosine kinase. Annu. Rev.Biochem. 57:443-478.

53. Woolford, J. W., A. McAuliffe, and L. R. Rohrschneider. 1988. Activation ofthe feline c-fins proto-oncogene: multiple alterations are required to generate a

fullytransformed phenotype. Cell. 55:965-977.

54.Roussel, M., J.Downing,C. W.Rettenmier,andC. J. Sherr. 1988. A point mutation in the extracellular domain ofthe human CSF-Ireceptor(c-fins proto-oncogeneproduct)activatesitstransforming potential.Cell. 55:978-988.

55.Furitsu,T.,H.Saito,A.M.Dvorak, L. B.Schwartz, A-M.A.Irani,J. F. Burdick, K.Ishizaka,andT.Ishizaka. 1989.Developmentofhumanmastcells in vitro. Proc.Natl.Acad.Sci. USA. 86:10039-10043.

56.Valent, P., E.Spanblochl, W. R.Sperr,C.Sillaber,K. M.Zsebo, H.Agis,

H. Strobl, K.Geissler, P. Bettelheim, and K. Lechner.1992. Induction of differ-entiationof human mast cells from bone marrow andperipheralblood

mononu-clearcellsby recombinant humanstemcellfactor/kit-ligandin long-term

cul-ture.Blood. 80:2237-2245.

57. von Ruden, T., S. Kandels, T.Radaszkiewicz, A. Ullrich, and E. F. Wagner. 1992. Developmentof a lethal mast cell disease in mice reconstituted with bone marrow cellsexpressingthev-erbBoncogene.Blood.79:3145-3158.

58.Sekido,Y.,Y.Obata,R.Ueda,T.Hida,M.Suyama, K.Shimokata,Y. Ariyoshi, andT.Takahashi. 1991.Preferential expression ofc-kit protooncogene

transcriptsin small cell lungcancer.Cancer Res. 51:2416-2419.

59.Strohmeyer, T., S. Peter, M. Hartmann, S.Munemitsu,R.Ackermann, A.

Ullrich,and D. J. Slamon. 1991.Expressionofthe hst- 1 and c-kit protooncogenes in humantesticulargermcell tumors.Cancer Res. 51:1811-1816.

60. Hanks, S. K.,A.M.Quinn, and T. Hunter. 1988. The protein kinase