exclusively5555tk.com

REVIEWAnalysisofthegenometopersonalizetherapyformelanomaMADavies1,2andYSamuels31DepartmentofMelanomaMedicalOncology,TheUniversityofTexasMDAndersonCancerCenter,Houston,TX,USA;2DepartmentofSystemsBiology,TheUniversityofTexasMDAndersonCancerCenter,Houston,TX,USAand3CancerGeneticsBranch,NationalHumanGenomeResearchInstitute,NationalInstitutesofHealth,Bethesda,MD,USAThetreatmentofcancerisbeingrevolutionizedbyanimprovedunderstandingofthegeneticeventsthatoccurintumors.
Advancesintheunderstandingoftheprevalenceandpatternsofmutationsinmelanomahaverecentlyledtoimpressiveresultsintrialsofpersonalized,targetedtherapiesforthisdisease.
Inthisreview,wewilldiscussthemoleculartargetsthathavebeenvalidatedclinically,additionalgeneticeventsthatarecandidatesforfuturetrials,andthechallengesthatremaintoimproveoutcomesfurtherinthisaggressivedisease.
Oncogene(2010)29,5545–5555;doi:10.
1038/onc.
2010.
323;publishedonline9August2010Keywords:melanoma;BRAF;NRAS;C-KIT;ERBB4;GNaQIntroductionItisestimatedthat68720patientswerediagnosedwithmelanoma,and8650patientsdiedofthisdisease,in2009intheUSA(Jemaletal.
,2009).
Melanomaislikelytobecomeanincreasinglyimportantpublichealthissue,asareviewoftheSEERdatabasefrom1950to2000foundanincreasein619%intheannualincidenceand165%inannualmortalityfromthisdisease,morethananyothercancer(Tsaoetal.
,2004a).
Interferon-a-2bistheonlyadjuvanttherapyapprovedbytheFoodandDrugAdministration(FDA)forpatientswithhigh-risklocallyadvancedmelanoma.
Thistreatmentisgivenfor12monthsandB90%ofpatientsexperiencesignicanttoxicities(Hauschildetal.
,2008).
AdjuvantInterferon-a-2binitiallydemonstratedasignicantsurvivalbenetversusplacebointheECOGtrialE1684(Kirkwoodetal.
,1996).
However,asubsequentanalysisafteraddi-tionalfollow-up,andapooledanalysisincludingthreeadditionallargeadjuvantInterferon-a-2btrials,failedtoshowasignicantimprovementinsurvivalversuscontrols(Kirkwoodetal.
,2004).
Forpatientswithmetastaticmelanoma,theonlyFDA-approvedtherapiesaredacarbazineandhigh-doseinterleukin-2(HD-IL2).
Dacarbazine,whichisacytotoxicchemotherapyagent,hasaclinicalresponserateofB10%,andalmostallresponsesarebrief.
HD-IL2,whichisabiologicalagentdesignedtostimulatetheimmunesystem,hasasimilarclinicalresponserateof16%(Atkinsetal.
,1999;Phanetal.
,2001).
Long-termdisease-freesurvivalhasbeenobservedinpatientswhoachieveacompleteresponsewithHD-IL2,butthisconstitutesonlyB6%ofthetreatedpatients(Atkinsetal.
,2000).
TheuseofHD-IL2islimitedbytheseveretoxicitiesofthisregimen,whichrequirespatientstobehospitalizedthroughoutadmin-istrationofthetreatment,andearlytrialshada2%treatment-relatedmortalityrate(Schwartzetal.
,2002).
Numeroustrialsofcombinatorialapproaches,includingmultiagentchemotherapyandbiochemotherapyregi-mens,havefailedtoimproveoutcomesinpatients(Tsaoetal.
,2004a).
Themanagementofmanycancersisenteringintoanerainwhichtreatmentsarebeingusedtoinhibitpathwaysactivatedbymutationsinthetumors.
Thisapproach,whichhasbeentermed'targetedtherapy',hasdemonstratedremarkableclinicalbenetinseveralcancers.
Theuseofkinaseinhibitorsisnowthestandardofcareinpatientswithchronicmyelogenousleukemia,HER2/neu-ampliedbreastcancer,clear-cellrenalcellcarcinomaandgastrointestinalstromaltumors(GIST;Daviesetal.
,2006).
Althoughtheseresultshavebeenimpressive,thesuccessfuldevelopmentoftheseap-proachesdependedupontheidenticationofgeneticabnormalitiesinthesediseases.
Thereisgrowingevidencethatthemajorityofmelanomasharborgeneticchangesinkeyproteinkinasesignalingpathways(Figure1).
Afterinitialfailures,recentclinicaltrialsoftargetedtherapyagentshavedemonstratedpromisingactivityinmelanoma.
Wewillreviewthetargetsthathavebeenvalidatedclinicallyinmelanoma,anddiscussnewtargetsthatarecandidatesfortestinginthefuture.
ClinicallyvalidatedtargetsBRAFIn2002,investigatorsfromtheSangerInstitutepub-lishedtheidenticationofpointmutationsinBRAFintumorsandcelllines(Daviesetal.
,2002).
Althoughmutationswereidentiedin1–10%ofavarietyoftumortypes,includingcolon,lungandovariancancers,Received16April2010;revised13June2010;accepted15June2010;publishedonline9August2010Correspondence:DrYSamuels,CancerGeneticsBranch,NationalHumanGenomeResearchInstitute,NationalInstitutesofHealth,50SouthDrive,MSC8000,Building50,Room5140,Bethesda,MD20892-8000,USA.
E-mail:samuelsy@mail.
nih.
govOncogene(2010)29,5545–5555&2010MacmillanPublishersLimitedAllrightsreserved0950-9232/10www.
nature.
com/oncstrikinglyoverhalfofthetestedmelanomashadamutationintheBRAFgene.
ThehighprevalenceofBRAFmutationsinmelanomahasbeenvalidatedinmultiplestudies.
Arecentmeta-analysisofsequencingresultsforover2700samplesreportedamutationrateof65%inmelanomacelllinesand42%inunculturedcutaneousmelanomas(HockerandTsao,2007).
Muta-tionsinBRAFarethemostcommonsomaticmutationsinthisdisease.
TheBRAFproteinisaserine/threoninekinase.
BRAFisacomponentoftheRAS-RAF-MEK-MAPKpathway,whichisactivatedinmanycancers(Gray-Schopferetal.
,2007).
ActivationofthissignalingpathwayhadpreviouslybeenimplicatedinmelanomabytheidenticationofactivatingmutationsinNRAS,whichoccurin15–25%ofmelanomas(HockerandTsao,2007).
Over40differentpointmutationsinBRAFhavebeenidentiedinmelanoma.
Approximately90%oftheBRAFmutationsinmelanomaaffectasinglesite,thevalineatposition600(V600).
Themostcommonmutation,V600E(89%ofBRAFmutations),markedlyincreasestheinvitrocatalyticactivityofBRAF,andresultsinconstitutiveactivationofbothMEKandMAPK(Daviesetal.
,2002;Wanetal.
,2004).
TheBRAFV600EmutationsaremutuallyexclusivewithactivatingNRASmutations(Tsaoetal.
,2004b;Goeletal.
,2006;Haluskaetal.
,2006).
However,overlapofNRASmutationshasbeenobservedwithBRAFmutationsthatdonotincreaseBRAFcatalyticactivity(Wanetal.
,2004;Heidornetal.
,2010).
TheidenticationoffrequentactivatingBRAFmutationsinmelanomawasrapidlyfollowedbyfunctionaltesting.
KnockdownoftheV600EmutantBRAFinmelanomacelllinesbysmallinterferingRNAinhibitedMEKactivation,decreasedcellproliferationandinvasion,andinducedcelldeath(Hingoranietal.
,2003;Sumimotoetal.
,2004).
Sorafenibisasmallmoleculeinhibitorofwild-typeBRAF,V600EBRAF,CRAFandanumberoftyrosinekinasereceptors(Strumberg,2005).
Similartotheknockdownstudies,sorafenibinhibitedthegrowthandsurvivalofBRAF-mutanthumanmelanomacells,andslowedthegrowthofmelanomaxenografts(Karasaridesetal.
,2004).
DespitethesepromisingpreclinicalresultsandthehighprevalenceofBRAFV600Emutations,aPhaseIIsingle-agentstudyofsorafenibinmetastaticmelanomareportedonlyonepartialclinicalresponseamong34patients(Eisenetal.
,2006).
Asubsequentnonrando-mizedtrialofsorafenibincombinationwithpaclitaxelandcarboplatinshowedmuchmorepromisingactivity,witha26%clinicalresponserate,andtheachievementofdiseasecontrolinthemajorityofpatients(Flahertyetal.
,2008).
However,arandomizedPhaseIIItrialdemonstratedthatsorafenibdidnotincreasetheclinicalresponseordiseasecontrolrateachievedbythetwochemotherapyagentsalone(Hauschildetal.
,2009).
ThedisappointingclinicalresultswithsorafenibraisedthequestionofwhethermutantBRAFwasagoodtherapeutictargetinmelanoma.
TheBRAFV600Emutationispresentinupto80%ofbenignnevi,whichhavevirtuallynomalignantpotential(Pollocketal.
,2003;Yazdietal.
,2003).
Studiesinhumancelllines,zebrashandtransgenicmicealsodemonstratedthatexpressionoftheBRAFV600Emutationalonefailedtofullytransformmelanocytes(Michaloglouetal.
,2005;Pattonetal.
,2005;Dankortetal.
,2009).
Thus,itwaspossiblethatalthoughBRAFmutationsareprevalentinmelanoma,theywerenotessential.
ThevalidityofmutantBRAFasatherapeutictargethasnowbeendemonstratedbyclinicaltrialswithanew,mutant-specicBRAFinhibitor.
PLX4032(alsoknownasRO5185426)isasmallmoleculeinhibitorwithselectiveactivityagainstV600-mutantBRAF.
PLX4032inhibitsV600EBRAFcatalyticactivitywithanIC50of13nm,ascomparedwith160nmforwild-typeBRAF,and>1000nmforseveralotherrelatedkinases(CSK,SRC,focaladhesionkinase,KDR;Tsaietal.
,2008).
Similarly,theIC50fortheantiproliferativeeffectofPLX4032issignicantlylowerincellswithV600-mutantBRAF(0.
3–1.
7mm)versuscellswithwild-typeBRAF(>10mm).
InthedoseescalationportionofthePhaseIclinicaltrialmetastaticmelanoma,56%ofpatientswiththeBRAFV600Emutationhadapartialresponseandanadditional31%hadaminorresponse.
NoresponseswereseeninpatientswithoutaBRAFV600Emutation(Flahertyetal.
,2008).
Preliminaryresultsfromthedoseexpansioncohortofanadditional31patientswithBRAFV600Emutationsreportedthat70%achievedaclinicalresponse(Fisheretal.
,2010).
ThedramaticclinicalactivityofPLX4032isveryexciting.
However,thereremainchallengestooptimizethisnewtherapeuticapproach.
ManyofthepatientswhoinitiallyrespondedtoPLX4032havesubsequentlyprogressed,withearlydatademonstratingamediandurationofresponsetoPLX4032ofapproximatelyFigure1Kinasesignalingpathwaysinmelanoma.
ThemajorityofmelanomasharborsomaticmutationsintheRAS-RAF-MEK-MAPKorPI3K-AKTpathways.
Activatingmutationsareindicatedbyrippledredandbluecircles,withintensitydenotingrelativeprevalence.
Lossoffunctionmutationsarehighlightedinred.
RPTK,receptorproteintyrosinekinase(thatis,c-KIT,ERBB4).
TherapyformelanomaMADaviesandYSamuels5546Oncogene8monthsinthePhaseItrial(Fisheretal.
,2010).
Themechanismsthatcauseresistancearecurrentlyunknown.
Recently,anumberofstudieshavedemon-stratedthatBRAFinhibitors,includingPLX4032,activateMEKandMAPKinmelanomacelllineswithwild-typeBRAF,includingcelllineswithmutantNRAS(Halabanetal.
,2010;Hatzivassiliouetal.
,2010;Heidornetal.
,2010).
Thesestudiesshowedthatinhibitionofthecatalyticactivityofwild-typeBRAFpromotestheformationofacomplexbetweenBRAFandCRAF,whichincreasesCRAFcatalyticactivity.
TheformationofthiscomplexwasRAS-dependentintwoofthestudies,butnotinthethird.
ActivatedCRAFactivatedMEKandERK,andinsomeexperimentsproducedincreasedproliferationandmigrationofmelanomacells.
Althoughthesestudiesdidnotspeci-callydemonstratemechanismsofresistanceinBRAFV600Emutantcelllines,anotherstudyreportedthatseveralBRAF-mutantcelllinesselectedforinvitroresistancetotheBRAFinhibitorAZ628hadincreasedCRAFproteinexpression(Montagutetal.
,2008).
Interestingly,therecentstudiesinthecelllineswithwild-typeBRAFreporteddifferentcomplexesandpathwaysthatareactivatedbydifferentBRAFinhibi-tors.
ThissuggeststhatdifferentapproachesmaybeneededtoovercomeresistancetothedifferentBRAFinhibitorsbeingtestedinpatients.
IfallresistancemechanismssharecontinuedactivationoftheRAS-RAF-MEK-ERKsignalingpathwayasacommondownstreamevent,combinatorialapproachesthattargetothercomponentsofthispathway,suchasMEK,maybeeffective(Emeryetal.
,2009).
AlthoughmostresearchtothispointhasfocuseduponmelanomaswithBRAFV600Emutations,thereisevidencethatmutationsaffectingotherresidueshavemarkedlydifferentsignalingeffectsandmaybesensitivetoothertherapeuticapproaches.
AnumberoftheBRAFmutationsthathavebeenidentiedinmelano-ma,includingD594V,G596RandG466V,donotactivatethecatalyticactivityofBRAF,butdoresultinactivationofMEKandERK(Wanetal.
,2004).
StudieshavedemonstratedthatmelanomacelllineswiththesemutationsarecriticallydependentuponCRAF,andaresensitiveinvitroandinvivotosorafenib(Smalleyetal.
,2009).
Thus,despitetheinitialnegativeclinicalresults,itispossiblethatpan-RAFinhibitors,likesorafenibmaybenetcertainsubsetsofmelanomapatients.
Testingthishypothesiswillrequireanalyzingpatientsfortheserelativelyraremutations,andcollabora-tiveeffortstoaccumulateenoughpatientstobeabletodenitivelycharacterizeclinicalresponsesinthiscohort.
Theobservedeffectsofthemutant-specicBRAFinhibitorsincelllineswithoutBRAFmutationsstronglysupportthattheseagentsshouldnotbeusedinpatientswithwild-typeBRAF.
Itremainstobeseeniftheoptimalapproachforthesepatients,whichrepresentover50%ofmelanomas,canderivesimilarclinicalbenetfrominhibitionoftheRAS-RAF-MEK-ERKpathway.
Alternatively,thesetumorsmaydependupondifferentpathways,andthustheirtreatmentsshouldbedirectedagainstothertargets.
c-KITTraditionally,melanomashavebeencategorizedbytheanatomicalsitetheprimarytumorarisesfrom.
Althoughmostpatientshavetumorsarisingfromsun-exposedskin,melanomasalsoarisefromacralsurfaces(palms,soles),mucosalsurfaces(intestines,reproductivetract)andtheuveaoftheeye.
In2005,aseminalpaperbyDrBorisBastiandemonstratedthatmelanomasarisingfromcutaneous,acralandmucosalsiteshavemarkedlydifferentpatternsofDNAcopygainandloss(Curtinetal.
,2005).
Amongthecutaneoustumors,therewerealsosignicantdifferencesbetweentumorswithandwithoutevidenceofchronicsundamage.
Inparallel,theinvestigatorsdemonstratedthattheprevalenceofBRAFandNRASmutationsvariedmarkedlybetweentheanatomicallydenedgroups(Table1).
WhereasactivatingBRAFmutationswerehighlyprevalentincutaneoustumorswithoutchronicsundamage(59%),themutationratewasmuchlowerincutaneoustumorswithchronicsundamage(11%),acralmelanomas(23%)andmucosalmelanomas(11%).
OtherstudieshaveshownthatBRAFmutationsareessentiallyundetectableinuvealmelanomasbytraditionaltechnologies(Cohenetal.
,2003;Rimoldietal.
,2003).
Analysiswithmoresensitivetechno-logiessuggeststhatasubpopulationofcellsinuvealmelanomasmayharborBRAFmutations(Maatetal.
,2008).
ThelowerprevalenceofBRAFmutationsinthelesscommonmelanomasledtoinvestigationstoidentifyotherdrivermutations.
TheanalysisofDNAcopynumberchangesidentied4q12asaregionofchromo-somalgainintheacralandmucosalmelanomas,butnotinthecutaneoustumors(Curtinetal.
,2005).
Characterizationofcandidategenesinthisregionledtothendingoffocalcopynumbergainandmutationsinthec-KITgeneinmelanoma(Curtinetal.
,2006).
c-KITisatyrosinegrowthfactorreceptorandB80%ofGISTharboranactivatingmutation(Hirotaetal.
,1998).
Inmelanoma,c-KITmutationswereidentiedin17%Table1FrequencyofmutationsinkinasesignalingpathwaysinmelanomasubtypesCutaneousAcralMucosalUvealNotspeciedBRAF40–6015–203–5o1NRAS15–2510–155–15o1c-KIT(mutation)o2(CSD:2–17)10–2015–20o1c-KIT(amplication)0–7(CSD:6)25–3025–30o1GNaQo1(CSD:5)o1o145–50ERBB410–20PTEN10–30PIK3CA1–2AKT1/31–2Abbreviation:CSD,chronicsun-damagedcutaneous.
Eachcellshowsthepercentageoftumorswiththeindicatedmutationoramplication.
TherapyformelanomaMADaviesandYSamuels5547Oncogenechronicsun-damagedcutaneous,11%acraland21%mucosalmelanomas.
Incontrast,noKITpointmuta-tionsweredetectedinareasthatwerenotexposedtochronicsundamage,suchasthetrunkandtheback(Curtinetal.
,2006).
Inaddition,KITgeneamplicationwaspresentin6%ofchronicsun-damaged,7%ofacrallentiginousand8%ofmucosalmelanomas.
Subsequentstudiesinotherpanelsofmelanomashaveidentiedsimilarratesofc-KITalterationsinacralandmucosaltumors,butlowerrates(B2%)inchronicsun-damagedcutaneoustumors(Antonescuetal.
,2007;Beadlingetal.
,2008;Riveraetal.
,2008;Torres-Cabalaetal.
,2009;Handoliasetal.
,2010).
Invitrostudiesdemon-stratedthatpointmutationsinc-KITresultinconsti-tutiveactivationofthec-KITproteininmelanomacells,andtheactivationofdownstreamproliferativeandpro-survivalsignalingpathways(Jiangetal.
,2008;Ashidaetal.
,2009).
Theidenticationofactivatingc-KITmutationswassurprising,asanumberoflinesofresearchpreviouslyrejectedaroleforc-KITfunctioninmelanoma.
Lossofc-KITproteinexpressionhasbeenshowntocorrelatewithmelanomaprogression,andenforcedexpressionofc-KIT-inhibitedmelanomagrowthinvitroandinvivo(LassamandBickford,1992;Huangetal.
,1996).
ThreePhaseIIclinicaltrialsofthec-KITinhibitorimatinib,whichisFDA-approvedforGIST,reportedonlyoneclinicalresponseamong63totalpatients(Ugureletal.
,2005;Wymanetal.
,2006;Kimetal.
,2008).
However,thesetrialswereconductedbeforetheidenticationofc-KITaberrationsinraremelanomasubtypesand,thus,theylikelywereoverwhelminglycomposedofpatientswithcutaneousmelanomas.
Theonepatientwhodidrespondhadaprimaryacralmelanoma.
Therearenowseveralcasereportsofindividualmelanomapatientswithc-KITmutationswhohaveachieveddramaticclinicalresponsestoc-KITsmallmoleculeinhibitors,includingimatinib,sorafenibanddasatinib(Antonescuetal.
,2007;Hodietal.
,2008;Quintas-Cardamaetal.
,2008;Woodmanetal.
,2009).
Theseresultssupportapotentialbenetfortheseinhibitorsinmelanomapatientswithc-KITgeneticaberrations,whicharecurrentlybeingtestedinseveralclinicaltrials.
InGIST,mutationsinc-KIT,butnotproteinexpressionlevels,arepredictiveofresponsetoc-KITinhibitors.
Inmelanoma,imatinibresultedina50%clinicalresponserateamong10patientswithc-KITmutations,butinnoneofthe10patientswithc-KITamplicationofthewild-typegene(Fisheretal.
,2010).
Althoughsomeresponsestoc-KITinhibitorshavebeendurable,manyresponsesprogresswithinarelativelyshorttime(Woodmanetal.
,2009).
Inaddition,thec-KITmutationsassociatedwithimatinib-resistanceinGIST(exons13,17and18)aremoreprevalentinmelanoma(B15%)ascomparedwithGIST(B1%).
Thus,itremainstobeseenhoweffectivec-KITinhibitorswillbeinmelanoma.
However,theinitialdemonstrationofactivityinmelanomapatientswithc-KITmutationssupportsfurtherresearchtobuilduponthisactivity.
PotentialtargetsNRASTheRASfamilyofgenesismonomericguanosinetriphosphatasesthatarenormallyactivatedbyextra-cellularsignals,whichconvertthemtoaguanosinetriphosphate-boundactivestate.
ActivatingmutationsofdifferentRASisoforms(thatis,KRAS,HRASandNRAS)arecommoninmanytypesofcancer,includingpancreatic(Almogueraetal.
,1988),colon(Bosetal.
,1987;Forresteretal.
,1987),non-smallcelllung(Rodenhuisetal.
,1987),acutemyelogenousleukemiaandthyroidcancers(Bosetal.
,1985;Paduaetal.
,1985;Suarezetal.
,1988;Bos,1989).
ThesemutationslockRASintheguanosinetriphosphate-boundactivestate.
MutantRASactivatestheRAS-RAF-MEK-MAPKpathway,similartomutantBRAF.
However,researchinavarietyofcelltypeshasdemonstratedthatRASactivatesanumberofothereffectorpathways,includingthephosphatidylinositol3-kinases(PI3K)-AKT,RALGDSandPLCecascades(Downward,2003).
MutationsinNRAShavesincebeenidentiedinB15–20%ofmelanomaswith90%ofmutationslocalizingtocodon61(Broseetal.
,2002;Daviesetal.
,2002;HockerandTsao,2007).
NRASmutationsaremutuallyexclusivewithactivatingBRAFmutations,andtheirprevalenceisrelativelyconsistentacrossthenon-uvealanatomicalsubtypes(Curtinetal.
,2005).
InhibitingmutantRASproteinshasproventobeadauntingchallenge(Konstantinopoulosetal.
,2007).
Tobeactivated,RASmustrsttranslocatetotheplasmamembrane.
ThistranslocationisdependentuponthetransferofafarnesylisoprenoidgrouptoRAS,whichisnormallycatalyzedbyfarnesyltransfer-ase.
Anumberoffarnesyltransferaseinhibitorsweredevelopedtoinhibitthisprocess,andweresubsequentlyevaluatedinclinicaltrials.
Unfortunately,theseagentsfailedtodemonstratesignicantactivity.
SubsequentstudiesdemonstratedthatalthoughHRASdependsexclusivelyuponfarnesylationformembranerecruit-ment,bothKRASandNRASalternativelyundergogeranylgernylation,whichresultsinrecruitmenttotheplasmamembrane(Zhangetal.
,1997).
Inaddition,farnesyltransferasesmodifymorethan60proteins,thusmakingitchallengingtointerprettheroleofRASintheeffectsoffarnesyltransferaseinhibitors.
BecauseofthedifcultyofinhibitingRASdirectly,therapeuticstrategieshavefocuseduponinhibitingdownstreameffectorpathways.
PreclinicalstudieswithMEKinhibitors,suchasAZD6244,havedemonstratedgrowthinhibitioninmost,butnotall,NRAS-mutantmelanomacelllines(Dryetal.
,2010).
Studiesinseveralsystems,includingmelanoma,haveshownincreasedefcacywithcombinedinhibitionoftargetsinmultipleRASeffectorpathways,suchasMEKandPI3K(Engelmanetal.
,2008;Jaiswaletal.
,2009b).
Atthistime,theefcacyoftargetedtherapiesinmelanomapatientswithNRASmutationsisyettobereported,butthiswillbeanimportantareaofstudyinthefuture.
TherapyformelanomaMADaviesandYSamuels5548OncogenePI3K-AKTActivatedNRASisknownnotonlytoactivatetheRAF-MEK-ERKpathway,butalsothePI3K-AKTpathway(Stahletal.
,2004;Tsaoetal.
,2004a,b).
PI3Ksareafamilyoflipidkinasesthatphosphorylatethe30-OHgroupoftheinositolring(Katsoetal.
,2001;Cantley,2002;VivancoandSawyers,2002).
ClassIAPI3Ksareheterodimerscomprisedofacatalyticsubunitencodedbythreep110genes(p110a,p110bandp110d)andvedifferentregulatorysubunits(p85a,p85b,p85g,p50aandp55a).
TheenzymaticproductgeneratedbythePI3Ks,PI-(3,4,5)P3,activatesAKT,aserine/threoninekinasethatinducesavarietyofresponsesincludingincreasedcellgrowthandproliferation(Katsoetal.
,2001;Cantley,2002;VivancoandSawyers,2002).
Phosphatidylinositol(3,4,5)triphosphatelevelsaretightlycontrolledbystrictregulationofPI3Kactivityandbyphosphataseandtensinhomolog(PTEN),aphosphatasewhichdephosphorylatesthe30-positionoftheinositolring(MaehamaandDixon,1998;Myersetal.
,1998).
PTENisatumorsuppressorgenethatiscommonlyinactivatedinhumantumors(Lietal.
,1997;Stecketal.
,1997),resultinginconstitutiveactivationofthePI3Kpathway.
SomaticinactivationofPTENisfoundin10–30%ofmelanomas(Guldbergetal.
,1997;Robertsonetal.
,1998;Tsaoetal.
,1998;Bircketal.
,2000;Zhouetal.
,2000;Chudnovskyetal.
,2005).
PTENlossusuallyoccursinmelanomasthatalsoharboraBRAFmutation,butitismutuallyexclusivewithNRASmutation.
Thus,melanomasfrequentlyhavegeneticactivationofboththePI3KandERKsignalingpathways,eitherbycombinedBRAFmutationandPTENloss,orbyNRASmutation(Tsaoetal.
,2004b).
However,quantitativemeasurementofactiva-tion-specicmarkershasdemonstratedthatPTENlossandNRASmutationsmaynothaveequivalenteffectsonAKTactivation(Daviesetal.
,2009).
RareactivatingmutationsofthecatalyticsubunitofPI3K,PIK3CA,havealsobeenreportedinmelanoma,butthemuta-tionalstatusofBRAFandNRASinthosetumorswasnotreported(Curtinetal.
,2006;Omholtetal.
,2006).
AlthoughactivationofthePI3K-AKTpathwayisfrequentlyimplicatedinoncogenesis,thereisevidencethatsomemediatorsofthispathwaymaybedifferentinmelanomathaninothercancers.
TherearethreeAKTisoforms,AKT1,2and3.
Theseisoformshaveverysimilarsequencesandstructures(Braziletal.
,2002).
AlthoughAKT1andAKT2arethoughttobethemostcriticalisoformsinmostcancers,thereareseveralpiecesofevidencesupportingacentralroleforAKT3inmelanoma.
First,bothimmunoprecipitationandsmallinterferingRNA-mediatedknockdownofthedifferentAKTisoformsdemonstratedthatAKT3isfrequentlythepredominantactive/phosphorylatedforminmela-nomacells(Stahletal.
,2004).
AKT3islocatedonthelongarmofchromosome1,whichisanareaoffrequentchromosomalgaininmelanoma(Thompsonetal.
,1995;Bastianetal.
,1998).
Recently,ararepointmutation(E17K)intheregulatorypleckstrinhomologydomainofAKT3wasidentiedinmelanomacelllinesandclinicalspecimens(Daviesetal.
,2009).
ThismutationisinthesameresidueasthepointmutationpreviouslyidentiedinAKT1inothercancertypes,includingbreast,ovaryandcoloncancers(Carptenetal.
,2007).
SimilartoAKT1mutations,theAKT3E17KmutationresultsinconstitutiveactivationofAKT.
AlthoughactivationofthePI3K-AKTpathwayappearstobeafrequenteventinmelanoma,todateclinicaltrialsusingagentsagainstthispathwayhavefailedtodemonstratedsignicantefcacy.
TreatmentwithCCI-779,ananalogofrapamycinthatinhibitstheAKTeffectormammaliantargetofrapamycin,yieldedonlya3%responserate(Margolinetal.
,2005).
Althoughmammaliantargetofrapamycininhibitorswereanappealingclinicalstrategyinitiallyowingtotheknowntoleranceforrapamycininpatients,researchhasdemonstratedthattheseagentsdisruptafeedbackloop,resultinginhyperactivationofPI3KandAKT(O'Reillyetal.
,2006;Taberneroetal.
,2008).
Perifosineisasmallmoleculeinhibitorthatinterfereswiththerecruitmentofmoleculeswithapleckstrinhomologydomain,suchasAKT,totheplasmamembrane.
Treatmentof18metastaticmelanomapatientswithperifosinefailedtoyieldanyclinicalresponses(Ernstetal.
,2005).
Unfortunately,noanalyseswereperformedinthistrialtoseeifperifosinetreatmentinhibitedAKTactivationinpatients.
ManynewagentsthattargetthePI3K-AKTpathway,includingPI3K,dualPI3K/mammaliantargetofrapamycinandcatalyticdomainAKTinhibitors,arenowinvariousphasesofclinicaltesting(Courtneyetal.
,2010).
Itremainstobeseenwhatactivitythesedrugswillhaveinmelanomaassingleagents.
However,thegeneticsofmelanomasuggeststhatsuchagentsmaybemosteffectivewhencombinedwithRAS-RAF-MEK-MAPKpathwayinhibitors.
ERBB4Althoughtheaforementionedtargetshavegeneratedgreatinterestastherapeutictargets,thereremainasignicantnumber(B30%)ofmelanomapatientsforwhomnoactivatingmutationisidentied.
Toidentifynewtargetsthatmaycontributetothisdisease,werecentlyinvestigated86proteintyrosinekinasesforsequencevariationsinasetofcutaneousmelanomas(Prickettetal.
,2009).
Afterinitiallyscreeningformutationsinthekinasedomainsonly,weperformedwhole-exomesequencingforthegeneswithasomaticmutationinthekinasedomain.
Usingthisapproach,weidentiedatotalof99nonsynonymoussomaticmuta-tionsin19proteintyrosinekinasegenes(Table2).
Althoughanumberofthesegenesarepotentialtherapeutictargets,asaninitialcandidatewefocusedontheERBB4gene.
ERBB4isamemberoftheERBB/HERfamilyofreceptortyrosinekinases.
Otherfamilymembers,includingERBB1(epidermalgrowthfactorreceptor)andERBB2(HER2),havebeenimplicatedbymuta-tionsand/oramplicationsinanumberofcancers,includinglung,colonandbreastcancers.
Geneticabnormalitiesinthesegenesareextremelyrareinmelanoma(Inmanetal.
,2003;Akslenetal.
,2008).
TherapyformelanomaMADaviesandYSamuels5549OncogeneConsistentwiththeseresults,wedidnotdetectanysomaticmutationsinthekinasedomainofERBB1/2/3inmelanomasamples.
However,theERBB4geneharboredmoresomaticchangesthaninanyotherproteintyrosinekinase.
Asmutationsthatariseduringtumorigenesismayprovideaselectiveadvantagetothetumorcell(drivermutations)orhavenofunctionaleffectontumorgrowth(passengermutations),itisimportanttocaptureboththenonsynonymousandthesynonymousalterationsforfurtheranalysis.
Theratiobetweenthesetwocategoriesofmutationsmayindicatethelikelihoodthatagenewasselectedforduringtumorigenesisandthereforeisadrivermutation.
InthecaseofERBB4,thisratiowas24:3,whichissignicantlyhigherthantheNS/Sratiopredictedfornonselectedpassengermutations(P>0.
01;Sjoblometal.
,2006).
ThissupportsthatthefrequentERBB4mutationswerenotrandomevents,butmaybefunctional,drivermutationsthatcontributetomelanomadevelopmentorprogression.
TherewasoverlapwithbothBRAFandNRASinthetumorswithERBB4mutations,suggestingthatthesegenesmayoperatethroughindependentpathways.
Indeed,thissamescenarioisseenforactivatingmutationsinothergenes,suchasPIK3CA(Samuelsetal.
,2004).
TheERBB4mutationswerelocalizedinseveraldifferentfunctionaldomainsoftheprotein.
Onthebasisofbioinformatics,analysesthatlocalizedtheidentiedmutationsontheproteincrystalstructureandevaluatedtheirproximitytothepreviouslyidenti-edmutationsinERBB1andERBB2,sevenofthemutationswereclonedintoexpressionvectorsforfunctionalexperiments.
Althoughthesesevenmutationsaffectedseveraldifferentfunctionaldomains,allofthemutationsinducedincreasedERBB4catalyticactivity(asmeasuredbyautophosphorylation),kinaseactivity(measuredusinganinvitrokinaseassay)andanchorage-independentgrowth.
Inaddition,allthemutationswereshowntobetransforming,toasimilarextent,tothewell-knownoncogeneKRASG12V.
WealsoexaminedtheeffectsofinhibitingERBB4inmelanomacelllineswiththegenemutations.
KnockdownofERBB4expressionwithsmallhairpinRNAhadminimaleffectsontheproliferationofhumanmelanomacelllineswithwild-typeERBB4.
Incontrast,celllineswithvariousERBB4mutationsweremarkedlyinhibitedbyknockdownofthegene.
Asimilarpatternofselectivesensitivitywasobservedwhenthecelllinesweretreatedwithlapatinib,whichisanFDA-approvedpan-ERBBfamilyinhibitor.
AlthoughlapatinibhasbeenshowntohavegreatestinhibitoryactivityagainstERBB1andERBB2,itshowedselectivegrowthinhibitioninthemelanomacelllinesexpressingmutantERBB4.
Themechanismthatliesbehindthisselectivityisasyetunknown,butitcouldbebecauseoftheinhibitionofthemutantERBB4proteinitself,oralternativelytoapreferentialheterodimerizationofmutantERBB4withERBB2.
AsummaryofthegeneticandfunctionalinvestigationofERBB4isdepictedinFigure2.
Takentogether,thesendingshaveidentiedanovelmelano-ma'driver'thatcauses'oncogeneaddiction'allowingfortheinvestigationofapplyingtargetedtherapeuticsformelanomapatientsharboringERBB4mutations.
InterestinglyERBB4mutationsaredispersedthro-ughoutitsdomains.
ThisisreminiscenttothemutationsreportedinPIK3CA;althoughithastwomajormutationalhotspots,itharborsmutationsthroughoutitsdomains(Samuelsetal.
,2004;ZhaoandVogt,2008).
Asimilarobservationisseenforp85a,whoseidentiedmutationsmainlylieintheiSH2domain,butarealsofoundinthecSH2,nSH2,iSH2cSHandBCRdomains,(Philpetal.
,2001;TCGA,2008;Jaiswaletal.
,2009a),aswellasFLT3(http://www.
sanger.
ac.
uk/perl/genetics/CGP/cosmicactiongene&lnFLT3).
Thissuggeststhatnotallmutationsinoncogenesmustbeclusteredtobefunctionallyimportant.
Aschangesthataffectpro-teinactivitycanresultfromsingleormultiplemutationswithinagenethatincreaseactivityorabrogatenegativeregulatorydomains.
GNaQAsdescribedpreviously,uvealmelanomasarenotableforanalmostcompleteabsenceofBRAFandNRASmutationsbyconventionalsequencingapproaches(Cruzetal.
,2003).
PreviousexperimentsusingaforwardgeneticscreenhadidentiedthathypermorphicmutationsintwodifferentG-protein-coupledreceptors,GNaQandGNa11,producedincreasedproliferationofdermalmelanocytes(VanRaamsdonketal.
,2009).
Asconditionsthatarecharacterizedbyincreasedprolifera-tionofintradermalmelanocytes(thatis,nevusofOta)areassociatedwithanincreasedriskofuvealmelanoma,investigatorsscreeneduvealmelanomasformutationsinTable2SomaticmutationsidentiedintheproteintyrosinekinasescreenGeneNo.
ofnonsynonymousmutationsfoundNo.
ofsynonymousmutationsfoundMelanomacasesaffected(%)aDDR1202.
5FER202.
5FLT18010.
1EPHA6516.
3EPHA10706.
3EPHB1415.
1EPHB2728.
9EPHB6728.
9ERBB424319MATK101.
3MET303.
8NTRK1202.
5PDGFRA515.
1PTK2111.
3PTK2B8110.
1PTK6202.
5PTK7111.
3ROR2415.
1TIE1637.
6Total9917aNumberofnonsynonymousandsplice-sitemutationsobservedandpercentageoftumorsaffectedforeachofthe19genesinthepanelof79cutaneousmelanomacancers.
TherapyformelanomaMADaviesandYSamuels5550Oncogenethesegenes.
TheteamleadbyDr.
BorisBastianidentiedarecurrentpointmutationintheGNaQgenein46%ofuvealmelanomasand27%ofuvealmelanomacelllines(VanRaamsdonketal.
,2009).
NoGNaQmutationswereidentiedinacral,mucosalorcutaneousmelanomaswithoutchronicsundamage;onecutaneoustumorwithchronicsundamageharboredamutation.
Astudybyanindependentgroupofinvesti-gatorssimilarlyreporteda49%incidenceofGNaQmutationsinprimaryuvealmelanomas(Onkenetal.
,2008).
AllthemutationsoccurredattheQ209residue,whichisanalogoustotheQ61residueinNRAS,andallweresomatic.
GNaQencodesforthea-subunitofqclassofheterotrimericguanosinetriphosphate-bindingprotein(Gq)thatmediatessignalsbetweenG-protein-coupledreceptors(GPCRs)andstimulatesphospholipaseC(PLC).
ActivationofPLCcatalyzesthehydrolysisofphosphatidylinositolbiphosphatetoreleaseinositoltrisphosphate(IP3)anddiacylglycerol(DAG).
ThesesecondmessengerspropagatetheGp-mediatedsignalthroughstimulationofproteinkinaseC(PKC).
ExpressionoftheGNaQQ209Lmutationpromotedanchorage-independentcellgrowthandtumorigenicityofmelanocytes(VanRaamsdonketal.
,2009).
Inaddition,expressionofGNaQQ209LactivatedproteinkinaseCsignalingandincreasedexpressionofP-ERK.
ApreliminaryreporthasalsodescribedpointmutationsinGNa11inuvealmelanomas,whichweremutuallyexclusivewiththeGNaQmutations(Fisheretal.
,2010).
FutureinvestigationofthispromisingtargetshouldencompassfurtherfunctionalassaystounderstandthesignalingmechanismsregulatedbymutantGNaQandGNa11.
Inaddition,directedtherapyagainstthemutantGNaQ/GNa11orthesignalingpathwaysFigure2GeneticandfunctionalanalysisofERBB4mutations.
Flowchartrepresentingtheexperimentaldesignusedtoevaluatethepresenceofsomaticmutationsinthetyrosinekinomeinmelanomaanditsfunctionalfollow-up.
OnceERBB4wasfoundtobethemosthighlymutatedtyrosinekinase,sevenofitsmutationswereshownto(fromlefttoright)increaseitskinaseactivity,inducecelltransformationandprovideanessentialcellsurvivalsignalasseeninsmallhairpinRNA(shRNA)knockdownandsmallmoleculeinhibitionassays.
WT,wildtype.
FigureadaptedfromPrickettetal.
,2009.
TherapyformelanomaMADaviesandYSamuels5551Oncogeneactivatedbythesemutationsmayopenupnewtherapeuticstrategiesformelanocytictumors,particu-larlyuvealtumors.
ThisnewdiscoveryagainpointstotheimportanceofthesuperfamilyofmonomericGproteinsandGPCRsincancer.
LookingtothefutureThepromisingresultswithPLX4032andc-KITinhibitorsinmelanomapatientswithBRAFandc-KITmutations,respectively,addtothegrowinglistofsuccessfulclinicalstrategiesexploitingactivatinggeneticeventsincancer.
Thesesuccesseshavegeneratedtremendousenthusiasmforthecontinueddevelopmentofpersonalized,targetedtherapyapproachesforthisaggressivedisease.
However,thereremainseveralcriticalgapsinourunderstandingoftheroleofsignalingpathwaysinmelanomathatwillbeessentialforimprovingoutcomesinthisdisease.
AlthoughthemajorityofpatientswithBRAFandc-KITmutationshaverespondedtoinhibitorsagainstthesetargets,bothprimaryandsecondaryresistancehavebeenobserved.
Itwillbeimportanttousebothpreclinicalmodelsandclinicalspecimenstoidentifythemechanismsthatunderliethisresistance.
Itispossiblethatsomepatientsmayonlyneedtreatmentwithasingleinhibitor.
However,thepatternofmutationsinmelanoma,suchasthefrequentco-occurrenceofBRAFmutationsandPTENloss,suggeststhatcombinatorialapproachesagainstmultiplepathwaysmaybemoreeffective.
TherecentidenticationofERBB4mutationsintumorswithbothBRAFandNRASmutationssuggestsothercombinatorialapproaches.
Althoughcandidateapproachesmaybedrivenbytheidentica-tionofgeneticchangesand/oractivatedpathways,analternativeapproachistheuseofbroad,unbiasedfunctionalscreenswithlibrariesofsmallmoleculesorsmallinterferingRNAs.
Suchapproachesmayidentifynovelgenesorpathwaysaseffectivecombinatorialtargets,andmayleadtofocusedinvestigationstounderstandtheroleoractivationofthosepathways.
Inaddition,theevidencethatimmunotherapies,suchasHD-IL2,occasionallyresultindurablecuressupportstherationaletoseeifcombinatorialapproacheswiththoseagentswillimprovethedurationoftheresponsesthatareachievedbytargetedtherapies.
Inanumberofdiseases,particularlyleukemias,thestemcellmodelhasbeenproposedasacriticalmechanismofthefailureofanticanceragents(BonnetandDick,1997;Baguley,2010).
Thestemcellmodelhypothesizesahierarchicalorganizationinwhichasmallsubpopulationofcells,thecancerstemcells,istumorigenic,canproliferateindenitely,andisulti-matelyresponsibleforthelong-termoutcomesofmanyanticancertherapies.
Thereisevidencethatmelanomadoesnottthismodel.
Usingamodiedxenograftmodelinwhichthenatural–killer-cellactivityoftheNOD/SCIDmiceusedtotesttumorigenicpotentialwasabrogated,itwasdemonstratedthatonaverage25%ofcellsfreshlyisolatedfrommelanomashavetumorigeniccapacity(Quintanaetal.
,2008).
Analysisof15differentmarkers,whichshowedheterogeneousexpressioninmelanomaclinicalspecimens,includingproteinsasso-ciatedwiththestem-cellphenotypeinothercancers,failedtodistinguishtumorigenicfromnontumorigeniccellsinthismodel.
Thus,atleastformelanomacells,tumorigeniccellsappeartobequiteprevalentintumors,notrare(Baker,2008;Eaves,2008).
Recently,asubpopulationofslow-cyclingcellsexpressingtheproteinJARID1Bhasbeenidentiedinbothmelanomatumorsandcelllines(Roeschetal.
,2010).
Experimentsdemonstratedthatexpressionofthisproteinwasnotaprerequisitefortumorigenicity,andindependentlyconrmedtheefcientformationoftumorsbysinglemelanomacellsinthemodiedxenotransplantationassay.
InadditiontoBRAFandc-KIT,thereisongoingresearchtoidentifytherapeuticstrategiestoapplytomelanomapatientswithNRAS,AKT,ERBB4,GNaQandGNa11mutations.
However,B30%ofmelanomapatientsdonothaveamutationdetectedinanyofthesegenes.
Thereisaclearneedforfocusedresearcheffortsonthislargecohortofpatients.
AsthecumulativeprevalenceofmutationsinBRAFandNRASissohighinthisdisease,itispossiblethatthesetumorsharborothermutationsthatactivatetheRAS-RAF-MEK-MAPKsignalingpathway.
Alternatively,thesetumorsmaybedrivenbyactivationofcompletelydifferentpathways,andthusrequiredistincttreatmentstrategies.
Newtechnologies,suchaswhole-exomecapture(Ngetal.
,2009;Turneretal.
,2009)andwhole-genomesequencing(Pleasanceetal.
,2010)maybeparticularlyinformativeinthesepatients.
Theseapproachescanrevealnotonlysingle-basealterationsincodingandregulatoryregions,butalsostructuralchangessuchasamplications,deletionsandtranslocations.
Databaseintegrationofthisinformationshouldrevealthemajorpathwaysthatdrivethesetumors.
Recently,therstpublicationtodescribetheresultsofwhole-genomesequencingofamelanomatumorhasbeenreported(Pleasanceetal.
,2010).
ThisanalysisidentiedB30000somaticalterations,whichisatleastoneorderofmagnitudehigherthanothersolidtumors.
Althoughmanyofthesechangeslikelyrepresentpassengereventsowingtoultravioletradiationexposure,mutationsresultinginover180nonsynonymousaminoacidsubstitutionsweredetected.
Thesenumberssuggestthatasubstantialnumberofmelanomaswillneedtobesequencedandcomparedtoidentifylikelydrivermutations,withstringentfunctionaltestingessentialforvalidationofnewtargets.
Insummary,thetreatmentofmelanomaisrapidlychangingowingtotheanalysisofthemelanomagenome.
Althoughsuchstudiestodatehavefocusedonthetreatmentofmetastatictumors,inthefuturethesemolecularinsightsmayleadtostrategiestoimprovetheearlydiagnosisofthisdisease,andperhapstostratifywhichpatientswithlocaltumorsrequireaggressiveinterventionstoreducetheriskofdiseaserecurrenceandspread.
Ultimately,improvedunder-standingofthemolecularbasisofthisdiseasemayevenTherapyformelanomaMADaviesandYSamuels5552Oncogeneleadtopreventativestrategies.
Thus,melanomaisanillnessinwhichfutureoutcomesinpatientsarelikelytobeintertwinedwithagrowthinourunderstandingofthegeneticeventsthatoccurinthisdisease.
ConictofinterestTheauthorsdeclarenoconictofinterest.
AcknowledgementsWethankDarrylLejaforgraphicalexpertizeandNatureGeneticsforpermissiontouseguresfrompaperdoi:10.
1038/ng.
438.
ThisworkwassupportedbygrantsfromASCO(YoungInvestigatorAward)andMDAndersonCancerCenterSPOREinMelanoma(DevelopmentalResearchGrant;5P50CA09345903PP-DRP3;toMAD)andtheIntramuralResearchProgramsoftheNationalHumanGenomeResearchInstitute,NationalInstitutesofHealth,USA(toYS).
ReferencesAkslenLA,PuntervollH,BachmannIM,StraumeO,VuhahulaE,KumarRetal.
(2008).
MutationanalysisoftheEGFR-NRAS-BRAFpathwayinmelanomasfromblackAfricansandothersubgroupsofcutaneousmelanoma.
MelanomaRes18:29–35.
AlmogueraC,ShibataD,ForresterK,MartinJ,ArnheimN,PeruchoM.
(1988).
Mosthumancarcinomasoftheexocrinepancreascontainmutantc-K-rasgenes.
Cell53:549–554.
AntonescuCR,BusamKJ,FranconeTD,WongGC,GuoT,AgaramNPetal.
(2007).
L576PKITmutationinanalmelanomascorrelateswithKITproteinexpressionandissensitivetospecickinaseinhibition.
IntJCancer121:257–264.
AshidaA,TakataM,MurataH,KidoK,SaidaT.
(2009).
PathologicalactivationofKITinmetastatictumorsofacralandmucosalmelanomas.
IntJCancer124:862–868.
AtkinsMB,KunkelL,SznolM,RosenbergSA.
(2000).
High-doserecombinantinterleukin-2therapyinpatientswithmetastaticmelano-ma:long-termsurvivalupdate.
CancerJSciAm6(Suppl1):S11–S14.
AtkinsMB,LotzeMT,DutcherJP,FisherRI,WeissG,MargolinKetal.
(1999).
High-doserecombinantinterleukin2therapyforpatientswithmetastaticmelanoma:analysisof270patientstreatedbetween1985and1993.
JClinOncol17:2105–2116.
BaguleyBC.
(2010).
Multidrugresistanceincancer.
MethodsMolBiol596:1–14.
BakerM.
(2008).
Melanomainmicecastsdoubtonscarcityofcancerstemcells.
Nature456:553.
BastianBC,LeBoitPE,HammH,BrockerEB,PinkelD.
(1998).
Chromosomalgainsandlossesinprimarycutaneousmelanomasdetectedbycomparativegenomichybridization.
CancerRes58:2170–2175.
BeadlingC,Jacobson-DunlopE,HodiFS,LeC,WarrickA,PattersonJetal.
(2008).
KITgenemutationsandcopynumberinmelanomasubtypes.
ClinCancerRes14:6821–6828.
BirckA,AhrenkielV,ZeuthenJ,Hou-JensenK,GuldbergP.
(2000).
MutationandalleliclossofthePTEN/MMAC1geneinprimaryandmetastaticmelanomabiopsies.
JInvestDermatol114:277–280.
BonnetD,DickJE.
(1997).
Humanacutemyeloidleukemiaisorganizedasahierarchythatoriginatesfromaprimitivehemato-poieticcell.
NatMed3:730–737.
BosJL.
(1989).
rasoncogenesinhumancancer:areview.
CancerRes49:4682–4689.
BosJL,FearonER,HamiltonSR,Verlaan-deVriesM,vanBoomJH,vanderEbAJetal.
(1987).
Prevalenceofrasgenemutationsinhumancolorectalcancers.
Nature327:293–297.
BosJL,ToksozD,MarshallCJ,Verlaan-deVriesM,VeenemanGH,vanderEbAJetal.
(1985).
Amino-acidsubstitutionsatcodon13oftheN-rasoncogeneinhumanacutemyeloidleukaemia.
Nature315:726–730.
BrazilDP,ParkJ,HemmingsBA.
(2002).
PKBbindingproteins.
GettinginontheAkt.
Cell111:293–303.
BroseMS,VolpeP,FeldmanM,KumarM,RishiI,GerreroRetal.
(2002).
BRAFandRASmutationsinhumanlungcancerandmelanoma.
CancerRes62:6997–7000.
CantleyLC.
(2002).
Thephosphoinositide3-kinasepathway.
Science296:1655–1657.
CarptenJD,FaberAL,HornC,DonohoGP,BriggsSL,RobbinsCMetal.
(2007).
AtransformingmutationinthepleckstrinhomologydomainofAKT1incancer.
Nature448:439–444.
ChudnovskyY,KhavariPA,AdamsAE.
(2005).
Melanomageneticsandthedevelopmentofrationaltherapeutics.
JClinInvest115:813–824.
CohenY,Goldenberg-CohenN,ParrellaP,ChowersI,MerbsSL,Pe'erJetal.
(2003).
LackofBRAFmutationinprimaryuvealmelanoma.
InvestOphthalmolVisSci44:2876–2878.
CourtneyKD,CorcoranRB,EngelmanJA.
(2010).
ThePI3Kpathwayasdrugtargetinhumancancer.
JClinOncol28:1075–1083.
CruzIIIF,RubinBP,WilsonD,TownA,SchroederA,HaleyAetal.
(2003).
AbsenceofBRAFandNRASmutationsinuvealmelanoma.
CancerRes63:5761–5766.
CurtinJA,BusamK,PinkelD,BastianBC.
(2006).
SomaticactivationofKITindistinctsubtypesofmelanoma.
JClinOncol24:4340–4346.
CurtinJA,FridlyandJ,KageshitaT,PatelHN,BusamKJ,KutznerHetal.
(2005).
Distinctsetsofgeneticalterationsinmelanoma.
NEnglJMed353:2135–2147.
DankortD,CurleyDP,CartlidgeRA,NelsonB,KarnezisAN,DamskyJrWEetal.
(2009).
Braf(V600E)cooperateswithPtenlosstoinducemetastaticmelanoma.
NatGenet41:544–552.
DaviesH,BignellGR,CoxC,StephensP,EdkinsS,CleggSetal.
(2002).
MutationsoftheBRAFgeneinhumancancer.
Nature417:949–954.
DaviesM,HennessyB,MillsGB.
(2006).
Pointmutationsofproteinkinasesandindividualisedcancertherapy.
ExpertOpinPharmacother7:2243–2261.
DaviesMA,Stemke-HaleK,LinE,TellezC,DengW,GopalYNetal.
(2009).
IntegratedmolecularandclinicalanalysisofAKTactivationinmetastaticmelanoma.
ClinCancerRes15:7538–7546.
DownwardJ.
(2003).
TargetingRASsignallingpathwaysincancertherapy.
NatRevCancer3:11–22.
DryJR,PaveyS,PratilasCA,HarbronC,RunswickS,HodgsonDetal.
(2010).
TranscriptionalpathwaysignaturespredictMEKaddictionandresponsetoselumetinib(AZD6244).
CancerRes70:2264–2273.
EavesCJ.
(2008).
Cancerstemcells:here,there,everywhereNature456:581–582.
EisenT,AhmadT,FlahertyKT,GoreM,KayeS,MaraisRetal.
(2006).
Sorafenibinadvancedmelanoma:aphaseIIrandomizeddiscontinuationtrialanalysis.
BrJCancer95:581–586.
EmeryCM,VijayendranKG,ZipserMC,SawyerAM,NiuL,KimJJetal.
(2009).
MEK1mutationsconferresistancetoMEKandB-RAFinhibition.
ProcNatlAcadSciUSA106:20411–20416.
EngelmanJA,ChenL,TanX,CrosbyK,GuimaraesAR,UpadhyayRetal.
(2008).
EffectiveuseofPI3KandMEKinhibitorstotreatmutantKrasG12DandPIK3CAH1047Rmurinelungcancers.
NatMed14:1351–1356.
ErnstDS,EisenhauerE,WainmanN,DavisM,LohmannR,BaetzTetal.
(2005).
PhaseIIstudyofperifosineinpreviouslyuntreatedpatientswithmetastaticmelanoma.
InvestNewDrugs23:569–575.
TherapyformelanomaMADaviesandYSamuels5553OncogeneFisherDE,BarnhillR,HodiFS,HerlynM,MerlinoG,MedranoEetal.
(2010).
Melanomafrombenchtobedside:meetingreportfromthe6thinternationalmelanomacongress.
PigmentCellMelanomaRes23:14–26.
FlahertyKT,SchillerJ,SchuchterLM,LiuG,TuvesonDA,RedlingerMetal.
(2008).
AphaseItrialoftheoral,multikinaseinhibitorsorafenibincombinationwithcarboplatinandpaclitaxel.
ClinCancerRes14:4836–4842.
ForresterK,AlmogueraC,HanK,GrizzleWE,PeruchoM.
(1987).
DetectionofhighincidenceofK-rasoncogenesduringhumancolontumorigenesis.
Nature327:298–303.
GoelVK,LazarAJ,WarnekeCL,RedstonMS,HaluskaFG.
(2006).
ExaminationofmutationsinBRAF,NRAS,andPTENinprimarycutaneousmelanoma.
JInvestDermatol126:154–160.
Gray-SchopferV,WellbrockC,MaraisR.
(2007).
Melanomabiologyandnewtargetedtherapy.
Nature445:851–857.
GuldbergP,thorStratenP,BirckA,AhrenkielV,KirkinAF,ZeuthenJ.
(1997).
DisruptionoftheMMAC1/PTENgenebydeletionormutationisafrequenteventinmalignantmelanoma.
CancerRes57:3660–3663.
HalabanR,ZhangW,BacchiocchiA,ChengE,ParisiF,AriyanSetal.
(2010).
PLX4032,aselectiveBRAFV600Ekinaseinhibitor,activatestheERKpathwayandenhancescellmigrationandproliferationofBRAFWTmelanomacells.
PigmentCellMelanomaRes23:190–200.
HaluskaFG,TsaoH,WuH,HaluskaFS,LazarA,GoelV.
(2006).
Geneticalterationsinsignalingpathwaysinmelanoma.
ClinCancerRes12:2301s–2307s.
HandoliasD,SalemiR,MurrayW,TanA,LiuW,VirosAetal.
(2010).
MutationsinKIToccuratlowfrequencyinmelanomasarisingfromanatomicalsitesassociatedwithchronicandinter-mittentsunexposure.
PigmentCellMelanomaRes23:210–215.
HatzivassiliouG,SongK,YenI,BrandhuberBJ,AndersonDJ,AlvaradoRetal.
(2010).
RAFinhibitorsprimewild-typeRAFtoactivatetheMAPKpathwayandenhancegrowth.
Nature464:431–435.
HauschildA,AgarwalaSS,TrefzerU,HoggD,RobertC,HerseyPetal.
(2009).
ResultsofaphaseIII,randomized,placebo-controlledstudyofsorafenibincombinationwithcarboplatinandpaclitaxelassecond-linetreatmentinpatientswithunresectablestageIIIorstageIVmelanoma.
JClinOncol27:2823–2830.
HauschildA,GogasH,TarhiniA,MiddletonMR,TestoriA,DrenoBetal.
(2008).
Practicalguidelinesforthemanagementofinterferon-alpha-2bsideeffectsinpatientsreceivingadjuvanttreatmentformelanoma.
Cancer112:982–994.
HeidornSJ,MilagreC,WhittakerS,NourryA,Niculescu-DuvasI,DhomenNetal.
(2010).
Kinase-deadBRAFandoncogenicRAScooperatetodrivetumorprogressionthroughCRAF.
Cell140:209–221.
HingoraniSR,JacobetzMA,RobertsonGP,HerlynM,TuvesonDA.
(2003).
SuppressionofBRAF(V599E)inhumanmelanomaabrogatestransformation.
CancerRes63:5198–5202.
HirotaS,IsozakiK,MoriyamaY,HashimotoK,NishidaT,IshiguroSetal.
(1998).
Gain-of-functionmutationsofc-kitinhumangastrointestinalstromaltumors.
Science279:577–580.
HockerT,TsaoH.
(2007).
Ultravioletradiationandmelanoma:asystematicreviewandanalysisofreportedsequencevariants.
HumMutat28:578–588.
HodiFS,FriedlanderP,CorlessCL,HeinrichMC,MacRaeS,KruseAetal.
(2008).
MajorresponsetoimatinibmesylateinKIT-mutatedmelanoma.
JClinOncol26:2046–2051.
HuangS,LucaM,GutmanM,McConkeyDJ,LangleyKE,LymanSDetal.
(1996).
Enforcedc-KITexpressionrendershighlymetastatichumanmelanomacellssusceptibletostemcellfactor-inducedapoptosisandinhibitstheirtumorigenicandmetastaticpotential.
Oncogene13:2339–2347.
InmanJL,KuteT,WhiteW,PettenatiM,LevineEA.
(2003).
AbsenceofHER2overexpressioninmetastaticmalignantmelano-ma.
JSurgOncol84:82–88.
JaiswalBS,JanakiramanV,KljavinNM,ChaudhuriS,SternHM,WangWetal.
(2009a).
Somaticmutationsinp85alphapromotetumorigenesisthroughclassIAPI3Kactivation.
CancerCell16:463–474.
JaiswalBS,JanakiramanV,KljavinNM,Eastham-AndersonJ,CuppJE,LiangYetal.
(2009b).
CombinedtargetingofBRAFandCRAForBRAFandPI3KeffectorpathwaysisrequiredforefcacyinNRASmutanttumors.
PLoSOne4:e5717.
JemalA,SiegelR,WardE,HaoY,XuJ,ThunMJ.
(2009).
Cancerstatistics,2009.
CACancerJClin59:225–249.
JiangX,ZhouJ,YuenNK,CorlessCL,HeinrichMC,FletcherJAetal.
(2008).
ImatinibtargetingofKIT-mutantoncoproteininmelanoma.
ClinCancerRes14:7726–7732.
KarasaridesM,ChiloechesA,HaywardR,Niculescu-DuvazD,ScanlonI,FriedlosFetal.
(2004).
B-RAFisatherapeutictargetinmelanoma.
Oncogene23:6292–6298.
KatsoR,OkkenhaugK,AhmadiK,WhiteS,TimmsJ,WatereldMD.
(2001).
Cellularfunctionofphosphoinositide3-kinases:implicationsfordevelopment,homeostasis,andcancer.
AnnuRevCellDevBiol17:615–675.
KimKB,EtonO,DavisDW,FrazierML,McConkeyDJ,DiwanAHetal.
(2008).
PhaseIItrialofimatinibmesylateinpatientswithmetastaticmelanoma.
BrJCancer99:734–740.
KirkwoodJ,StrawdermanM,ErnstoffM,SmithT,BordenE,BlumR.
(1996).
Interferonalfa-2badjuvanttherapyofhigh-riskresectedcutaneousmelanoma:theEasternCooperativeOncologyGroupTrialEST1684.
JClinOncol14:7–17.
KirkwoodJM,ManolaJ,IbrahimJ,SondakV,ErnstoffMS,RaoU.
(2004).
Apooledanalysisofeasterncooperativeoncologygroupandintergrouptrialsofadjuvanthigh-doseinterferonformelano-ma.
ClinCancerRes10:1670–1677.
KonstantinopoulosPA,KaramouzisMV,PapavassiliouAG.
(2007).
Post-translationalmodicationsandregulationoftheRASsuper-familyofGTPasesasanticancertargets.
NatRevDrugDiscov6:541–555.
LassamN,BickfordS.
(1992).
Lossofc-kitexpressioninculturedmelanomacells.
Oncogene7:51–56.
LiJ,YenC,LiawD,PodsypaninaK,BoseS,WangSIetal.
(1997).
PTEN,aputativeproteintyrosinephosphatasegenemutatedinhumanbrain,breast,andprostatecancer.
Science275:1943–1947.
MaatW,KilicE,LuytenGPM,deKleinA,JagerMJ,GruisNAetal.
(2008).
PyrophosphorolysisdetectsB-RAFmutationsinprimaryuvealmelanoma.
InvestOphthalmolVisSci49:23–27.
MaehamaT,DixonJE.
(1998).
Thetumorsuppressor,PTEN/MMAC1,dephosphorylatesthelipidsecondmessenger,phospha-tidylinositol3,4,5-trisphosphate.
JBiolChem273:13375–13378.
MargolinK,LongmateJ,BarattaT,SynoldT,ChristensenS,WeberJetal.
(2005).
CCI-779inmetastaticmelanoma:aphaseIItrialoftheCaliforniaCancerConsortium.
Cancer104:1045–1048.
MichaloglouC,VredeveldLC,SoengasMS,DenoyelleC,KuilmanT,vanderHorstCMetal.
(2005).
BRAFE600-associatedsenescence-likecellcyclearrestofhumannaevi.
Nature436:720–724.
MontagutC,SharmaSV,ShiodaT,McDermottU,UlmanM,UlkusLEetal.
(2008).
ElevatedCRAFasapotentialmechanismofacquiredresistancetoBRAFinhibitioninmelanoma.
CancerRes68:4853–4861.
MyersMP,PassI,BattyIH,VanderKaayJ,StolarovJP,HemmingsBAetal.
(1998).
ThelipidphosphataseactivityofPTENiscriticalforitstumorsuppressorfunction.
ProcNatlAcadSciUSA95:13513–13518.
NgSB,TurnerEH,RobertsonPD,FlygareSD,BighamAW,LeeCetal.
(2009).
Targetedcaptureandmassivelyparallelsequencingof12humanexomes.
Nature461:272–276.
O'ReillyKE,RojoF,SheQB,SolitD,MillsGB,SmithDetal.
(2006).
mTORinhibitioninducesupstreamreceptortyrosinekinasesignalingandactivatesAkt.
CancerRes66:1500–1508.
OmholtK,KrockelD,RingborgU,HanssonJ.
(2006).
MutationsofPIK3CAarerareincutaneousmelanoma.
MelanomaRes16:197–200.
OnkenMD,WorleyLA,LongMD,DuanS,CouncilML,BowcockAMetal.
(2008).
OncogenicmutationsinGNAQoccurearlyinuvealmelanoma.
InvestOphthalmolVisSci49:5230–5234.
TherapyformelanomaMADaviesandYSamuels5554OncogenePaduaRA,BarrassNC,CurrieGA.
(1985).
ActivationofN-rasinahumanmelanomacellline.
MolCellBiol5:582–585.
PattonEE,WidlundHR,KutokJL,KopaniKR,AmatrudaJF,MurpheyRDetal.
(2005).
BRAFmutationsaresufcienttopromoteneviformationandcooperatewithp53inthegenesisofmelanoma.
CurrBiol15:249–254.
PhanGQ,AttiaP,SteinbergSM,WhiteDE,RosenbergSA.
(2001).
Factorsassociatedwithresponsetohigh-doseinterleukin-2inpatientswithmetastaticmelanoma.
JClinOncol19:3477–3482.
PhilpAJ,CampbellIG,LeetC,VincanE,RockmanSP,WhiteheadRHetal.
(2001).
Thephosphatidylinositol30-kinasep85alphageneisanoncogeneinhumanovarianandcolontumors.
CancerRes61:7426–7429.
PleasanceED,CheethamRK,StephensPJ,McBrideDJ,HumphraySJ,GreenmanCDetal.
(2010).
Acomprehensivecatalogueofsomaticmutationsfromahumancancergenome.
Nature463:191–196.
PollockPM,HarperUL,HansenKS,YudtLM,StarkM,RobbinsCMetal.
(2003).
HighfrequencyofBRAFmutationsinnevi.
NatGenet33:19–20.
PrickettTD,AgrawalNS,WeiX,YatesKE,LinJC,WunderlichJRetal.
(2009).
AnalysisofthetyrosinekinomeinmelanomarevealsrecurrentmutationsinERBB4.
NatGenet41:1127–1132.
QuintanaE,ShackletonM,SabelMS,FullenDR,JohnsonTM,MorrisonSJ.
(2008).
Efcienttumourformationbysinglehumanmelanomacells.
Nature456:593–598.
Quintas-CardamaA,LazarAJ,WoodmanSE,KimK,RossM,HwuP.
(2008).
CompleteresponseofstageIVanalmucosalmelanomaexpressingKITVal560Asptothemultikinaseinhibitorsorafenib.
NatClinPractOncol5:737–740.
RimoldiD,SalviS,LienardD,LejeuneFJ,SpeiserD,ZografosLetal.
(2003).
LackofBRAFmutationsinuvealmelanoma.
CancerRes63:5712–5715.
RiveraRS,NagatsukaH,GunduzM,CengizB,GunduzE,SiarCHetal.
(2008).
C-kitproteinexpressioncorrelatedwithactivatingmutationsinKITgeneinoralmucosalmelanoma.
VirchowsArch452:27–32.
RobertsonGP,FurnariFB,MieleME,GlendeningMJ,WelchDR,FountainJWetal.
(1998).
InvitrolossofheterozygositytargetsthePTEN/MMAC1geneinmelanoma.
ProcNatlAcadSciUSA95:9418–9423.
RodenhuisS,vandeWeteringML,MooiWJ,EversSG,vanZandwijkN,BosJL.
(1987).
MutationalactivationoftheK-rasoncogene.
Apossiblepathogeneticfactorinadenocarcinomaofthelung.
NEnglJMed317:929–935.
RoeschA,Fukunaga-KalabisM,SchmidtEC,ZabierowskiSE,BraffordPA,VulturAetal.
(2010).
Atemporarilydistinctsubpopulationofslow-cyclingmelanomacellsisrequiredforcontinuoustumorgrowth.
Cell141:583–594.
SamuelsY,WangZ,BardelliA,SillimanN,PtakJ,SzaboSetal.
(2004).
HighfrequencyofmutationsofthePIK3CAgeneinhumancancers.
Science304:554.
SchwartzRN,StoverL,DutcherJ.
(2002).
Managingtoxicitiesofhigh-doseinterleukin-2.
Oncology(WillistonPark)16:11–20.
SjoblomT,JonesS,WoodLD,ParsonsDW,LinJ,BarberTDetal.
(2006).
Theconsensuscodingsequencesofhumanbreastandcolorectalcancers.
Science314:268–274.
SmalleyKS,XiaoM,VillanuevaJ,NguyenTK,FlahertyKT,LetreroRetal.
(2009).
CRAFinhibitioninducesapoptosisinmelanomacellswithnon-V600EBRAFmutations.
Oncogene28:85–94.
StahlJM,SharmaA,CheungM,ZimmermanM,ChengJQ,BosenbergMWetal.
(2004).
DeregulatedAkt3activitypromotesdevelopmentofmalignantmelanoma.
CancerRes64:7002–7010.
SteckPA,PershouseMA,JasserSA,YungWK,LinH,LigonAHetal.
(1997).
Identicationofacandidatetumoursuppressorgene,MMAC1,atchromosome10q23.
3thatismutatedinmultipleadvancedcancers.
NatGenet15:356–362.
StrumbergD.
(2005).
Preclinicalandclinicaldevelopmentoftheoralmultikinaseinhibitorsorafenibincancertreatment.
DrugsToday(Barc)41:773–784.
SuarezHG,DuVillardJA,CaillouB,SchlumbergerM,TubianaM,ParmentierCetal.
(1988).
Detectionofactivatedrasoncogenesinhumanthyroidcarcinomas.
Oncogene2:403–406.
SumimotoH,MiyagishiM,MiyoshiH,YamagataS,ShimizuA,TairaKetal.
(2004).
InhibitionofgrowthandinvasiveabilityofmelanomabyinactivationofmutatedBRAFwithlentivirus-mediatedRNAinterference.
Oncogene23:6031–6039.
TaberneroJ,RojoF,CalvoE,BurrisH,JudsonI,HazellKetal.
(2008).
Dose-andschedule-dependentinhibitionofthemammaliantargetofrapamycinpathwaywitheverolimus:aphaseItumorpharmacodynamicstudyinpatientswithadvancedsolidtumors.
JClinOncol26:1603–1610.
TCGA(2008).
Comprehensivegenomiccharacterizationdeneshu-manglioblastomagenesandcorepathways.
Nature455:1061–1068.
ThompsonFH,EmersonJ,OlsonS,WeinsteinR,LeavittSA,LeongSPetal.
(1995).
Cytogeneticsof158patientswithregionalordisseminatedmelanoma.
Subsetanalysisofnear-diploidandsimplekaryotypes.
CancerGenetCytogenet83:93–104.
Torres-CabalaCA,WangWL,TrentJ,YangD,ChenS,GalbinceaJetal.
(2009).
CorrelationbetweenKITexpressionandKITmutationinmelanoma:astudyof173caseswithemphasisontheacral-lentiginous/mucosaltype.
ModPathol22:1446–1456.
TsaiJ,LeeJT,WangW,ZhangJ,ChoH,MamoSetal.
(2008).
DiscoveryofaselectiveinhibitorofoncogenicB-Rafkinasewithpotentantimelanomaactivity.
ProcNatlAcadSciUSA105:3041–3046.
TsaoH,AtkinsMB,SoberAJ.
(2004a).
Managementofcutaneousmelanoma.
NEnglJMed351:998–1012.
TsaoH,GoelV,WuH,YangG,HaluskaFG.
(2004b).
GeneticinteractionbetweenNRASandBRAFmutationsandPTEN//MMAC1inactivationinmelanoma.
JInvestDermatol122:337–341.
TsaoH,ZhangX,BenoitE,HaluskaFG.
(1998).
IdenticationofPTEN/MMAC1alterationsinunculturedmelanomasandmelanomacelllines.
Oncogene16:3397–3402.
TurnerEH,LeeC,NgSB,NickersonDA,ShendureJ.
(2009).
Massivelyparallelexoncaptureandlibrary-freeresequencingacross16genomes.
NatMethods6:315–316.
UgurelS,HildenbrandR,ZimpferA,LaRoseeP,PaschkaP,SuckerAetal.
(2005).
Lackofclinicalefcacyofimatinibinmetastaticmelanoma.
BrJCancer92:1398–1405.
VanRaamsdonkCD,BezrookoveV,GreenG,BauerJ,GauglerL,O'BrienJMetal.
(2009).
FrequentsomaticmutationsofGNAQinuvealmelanomaandbluenaevi.
Nature457:599–602.
VivancoI,SawyersCL.
(2002).
Thephosphatidylinositol3-kinaseAKTpathwayinhumancancer.
NatRevCancer2:489–501.
WanPTC,GarnettMJ,RoeSM,LeeS,Niculescu-DuvazD,GoodVMetal.
(2004).
MechanismofactivationoftheRAF-ERKsignalingpathwaybyoncogenicmutationsofB-RAF.
Cell116:855–867.
WoodmanSE,TrentJC,Stemke-HaleK,LazarAJ,PriclS,PavanGMetal.
(2009).
ActivityofdasatinibagainstL576PKITmutantmelanoma:molecular,cellular,andclinicalcorrelates.
MolCancerTher8:2079–2085.
WymanK,AtkinsMB,PrietoV,EtonO,McDermottDF,HubbardFetal.
(2006).
MulticenterphaseIItrialofhigh-doseimatinibmesylateinmetastaticmelanoma:signicanttoxicitywithnoclinicalefcacy.
Cancer106:2005–2011.
YazdiAS,PalmedoG,FlaigMJ,PuchtaU,ReckwerthA,RuttenAetal.
(2003).
MutationsoftheBRAFgeneinbenignandmalignantmelanocyticlesions.
JInvestDermatol121:1160–1162.
ZhangFL,KirschmeierP,CarrD,JamesL,BondRW,WangLetal.
(1997).
CharacterizationofHa-ras,N-ras,Ki-Ras4A,andKi-Ras4BasinvitrosubstratesforfarnesylproteintransferaseandgeranylgeranylproteintransferasetypeI.
JBiolChem272:10232–10239.
ZhaoL,VogtPK.
(2008).
ClassIPI3Kinoncogeniccellulartransformation.
Oncogene27:5486–5496.
ZhouXP,GimmO,HampelH,NiemannT,WalkerMJ,EngC.
(2000).
EpigeneticPTENsilencinginmalignantmelanomaswithoutPTENmutation.
AmJPathol157:1123–1128.
TherapyformelanomaMADaviesandYSamuels5555Oncogene