illustrationbyatelierolschinsky

When cells divide, chromosomes aren’t always doled out equally. The consequences can be dire.

bymeganscudellari

14 Fall 2014 / HHMI Bulletin

AAngelikaAmonhaswitnessedplentyofcelldivisionerrorsinhercareerasayeastgeneticist.Inyeast,theprocessofmeiosis—wheregermcellsdividetoformreproductivegametes—iserrorprone.Ascellsdivide,thread-likespindlefibersaresupposedtopullstubbychromosomepairsapartandsendthetwochromosomestooppositeendsofthecell.Sometimes,however,apairofchromosomesismistakenlydeliveredtoasingleside.Afterthecytoplasmdivides,theresultingtwodaughtercellsareaneuploid:onecarriesanextrachromosome;theotherismissingachromosome.

SittingatamicroscopetoexamineasetofaneuploidcellsinherlabattheMassachusettsInstituteofTechnology(MIT),Amonknewthecellswouldbesickly.Aneuploidyeastrarelyreplicate;theyoftendie.That’sbecauseorganisms—whethertheyhaveonesetofchromosomes(haploid,likesomeyeast),two(diploid,likemammals),or12(dodecaploid,liketheUgandaclawedfrog)—relyonbalancedgenomes.Anequalnumberofchromosomesassuresabalancedratioofgeneproducts.Whenthatbalanceisoutofwhack—whenanucleushasoneormoreextraormissingchromosomes—thecellalmostalwaysfailstodeveloporfunctionproperly.

Almostalways.Cancercells,whichareoverwhelminglyaneuploid,areanexception.Theythriveanddividewithafury.ThecontradictionnaggedAmon,anHHMIinvestigator,foryears.Howcouldaneuploidy—thekissofdeathformostcells—beahallmarkinthosecellswithanunparalleledabilitytoproliferate?Itdidn’tmakesense.

Whatwasworsewasthat,outsideofDownsyndrome(awell-studiedgeneticdisordercausedbyanextrachromosome21),aneuploidywasamystery.Despitetheprevalenceofaneuploidyincancercells—morethan90percentofsolidtumorsandabout75percentofblood

cancersdisplayit—scientistshadlittleunderstandingofhowaneuploidyaffectsthephysiologyofacell.

“Italwaysbuggedme,”saysAmon.“SoIdecidedtolookintoit.”Startingin2002,Amonsetouttodeterminethecellularimpactofhavinganabnormalnumberofchromosomesandtoresolvetheaneuploidyparadox.Beforelong,shehadalotofcompany.

Today,arosterofexperiencedcancerresearchersandmolecularbiologistsistacklinganeuploidy’seffectoncellsanditsroleincancer.Theireffortsarenotpurelyacademic:aneuploidycouldbeausefultargetfornewcancerdrugs.

Leaningforwardinachairinherfifth-storyoffice,Amonpropsherelbowsatopatable,firmlyclaspsherhands,anddeclares,“Iwanttoidentifygeneticorevenchemicalcompoundsthatkillaneuploidcellspreferentially.”Then,hervoicegrowingsoft,AmondescribeswatchingheryoungersisterwithbreastcancerperseverethroughnumerousroundsoftreatmentwithTaxol,acommonanti-cancerdrug.“It’sbrutal.Ifwecouldcombinepotentialaneuploidy-selectivecompoundswithTaxol,wecouldbemoreeffectiveandlowerthedoseofTaxol.Wecouldmakethelivesofpeoplewithcancersomuchbetter.”

Extra Gain, Extra PainAmon’sfirstforayintoaneuploidywasashotinthedark.“IhadatalentedtechniciannamedMonicaBoselli.SheandIusedtodocrazyexperimentsallthetime,”saysAmon.In2002,sheaskedBosellitomakeafewaneuploidcelllinessotheycouldgetagoodlookatthem—easiersaidthandone.First,Bosellilearnedanarcanechromosometransferstrategyusingmolecularmarkerstoselectandmoveindividualchromosomesfromoneyeastcelltoanother.Shecreated13strainsofhaploidyeast,eachwithoneortwoextrachromosomes.Then,Bosellicloselymonitoredthetransplantedchromosomes,asyeasthaveatalentforriddingthemselvesofextraneousgeneticmaterial.

Boselli’seffortsworked,andsheandAmon,alongwithAmon’sthenpostdocEduardoTorres,immediatelyobservedthattheextrachromosomewasnotdormant.Ineverycase,thechromosomewasactive:hundredsofgeneswerebeingexpressedfromitsDNA,and,Amonlaterconfirmed,thosegenetranscriptsweretranslatedintoproteins.

Next,Amonexaminedtheappearanceandactivityoftheaneuploidyeast.All13aneuploidcelllines,regardlessofwhichoftheyeast’s16chromosomeswasaddedastheextra,sharedcommontraits.Theyallhaddefectivecellcycles:theymultipliedmoreslowly—somecouldn’tevenformcolonies—andmanywere10-20minutesslowertoinitiatecellcycledivisionthantheirnormal,wild-typecounterparts.Thisconfirmedacommonassumptionthataneuploidyreducesacell’sabilitytoreproducenormally.Theyeastalsoexhibitedincreasedglucoseuptakeand

15HHMI Bulletin / Fall 2014A

ynsl

eyF

loyd

sensitivitytogrowthconditionsthathinderedproteinfoldingandthebreakdownofproteins.

Allinall,thefindingssuggestedthataneuploidcellsrequireextraenergyforsurvival,andthattheystruggletofoldanddegradeproteins.Theproteindegradationpathwaysappearedtobeworkingatfulltilttryingtoridthecellofunfoldedandmisfoldedproteinsandpartsofproteins.Thesubunitsofmanyproteincomplexesareencodedacrossmultiplechromosomes,soanextraormissingchromosomewillresultinisolatedsubunitswithnopartners.Infact,workpublishedthisAprilinCellbyHHMIInvestigatorJonathanWeissman,attheUniversityofCalifornia,SanFrancisco,showedthathealthycellsproduceequalamountsoftheproteinsthatfunctiontogetherinacomplex.Thus,anychangeinthosecarefullymaintainedratiosislikelytocauseproblemsinacell.

“Itwasprettyobvious[aneuploidcells]hadmajorissuesintheproteinqualitycontroldepartment,”saysAmon.She,Torres,andBosellipublishedtheresultsin2007.Afewyearslater,Amonfollowedupwithimagesofaneuploidyeastchokedwithclumpsoffluorescentlylabeledproteinaggregatesshininglikegreenwartsinthecells.

Amonlovesyeast:thesmellofyeast,thepinkandivorycolonies,theteenybubble-likeyeastspores.Butafter

onetoomanycolleaguesinsinuatedthatheraneuploidyfindingswereapplicableonlytoyeast,Amonturnedtomousecellstodetermineifwhattheyfoundinyeastwasalsotrueinmammals.

Assheexpected,themousecellshadcellcycledefectsandstressedproteincontrolpathways.“Itwasthesame,”saysAmonofthemammalianworkpublishedin2008inScience.“Definitelythesame.”ApaperpublishedthisyearbyZuzanaStorchovaandcolleaguesattheMaxPlanckInstituteofBiochemistryinGermanyconfirmsAmon’sfindings,thistimeinhumancells:aneuploidhumancells,fromavarietyofsources,upregulatemetabolicandproteindegradationpathways.

Tug of WarThreemilessouthandacrosstheCharlesRiver,DavidPellman,alsoanHHMIinvestigatorandyeastaficionado,wassimultaneouslytryingtofigureouthowaneuploidyaffectsthefunctionofcellsandobtainmoredetailsabouthowitoccursinthefirstplace.

Inadditiontoanabnormalnumberofchromosomes,mostaneuploidcellshaveanabnormalnumberofcentrosomes—smallorganellesthatcontrolthepositionandactionofchromosomesonspindlefibersduringcelldivision.Healthycellshavejusttwocentrosomesduringmitosis—divisionofsomatic,ornon-germ,cells—butaneuploidcellsoftenhavethreeorfour.In2008,attheDana-FarberCancerInstituteandBostonChildren’sHospital,Pellmanandcolleaguesbegangeneratingcelllines(avarietyofmouseandhuman)differingonlyincentrosomenumber.Atthetime,itwaswellestablishedthatextracentrosomesgenerateaneuploidy,butthemechanismbywhichthatoccurredhadnotbeendirectlyobserved.

Angelika Amon is searching for the link between aneuploidy and cancer.

“It was pretty obvious [aneuploidy cells] had major issues in the protein quality control department.”—angelikaamon

16 Fall 2014 / HHMI Bulletin

Ph

oto:

Am

yG

uti

erre

z

Watchingthecellsbyusingaliveimagingsystem,Pellman’steamfoundthatcellswithextracentrosomesdivideintotwo,justlikenormalcells,yettheextracentrosomesproduceadditionalspindlefibersthatpullchromosomesinthewrongdirections.Thistugofwarcanresultinlaggingchromosomesthatdon’tmakeittothecorrectsideofthecellpriortodivision.Thedaughtercells,therefore,areaneuploid.“Itwasanunexpectedmechanism,”saysPellman,butonethatconfirmedthatcentrosomesareactiveparticipantsinchromosomesegregationerrorsthatleadtoaneuploidy.

Aneuploidycanalsobecausedbyaweakenedordamagedspindlecheckpoint,thesecuritysysteminthecellthatensureschromosomesareaccuratelyseparated,saysHongtaoYu,anHHMIinvestigatorattheUniversityofTexasSouthwesternMedicalCenterinDallas.Ahostofspindlecheckpointproteins,suchasMad2andBubR1,monitortheattachmentofspindlefiberstokinetochores,stickyproteinhubsatthecenterofchromosomes.Inthecaseofincorrectfiber-kinetochoreattachments,checkpointproteinsbindandinhibitaproteincomplexthatwouldactivatethenextstageofcelldivision.

Completelossofthespindlecheckpointwhollykillscells,saysYu.Butlesserdefects,suchaslowlevelsordecreasedactivityofMad2orBubR1,resultinirregularattachments,continuedcelldivision,andchromosomemis-segregation.“Subtlechangesofthe

Hongtao Yu studies how subtle changes in spindle fiber proteins can cause chromosome mis-segregation.

Mitotic checkpoint defects

Cohesion defect

A

B

17HHMI Bulletin / Fall 2014D

iag

ram

:Ad

apte

dfr

omA

JHol

lan

da

nd

DW

Cle

vela

nd

,EM

BO

Rep

orts

13(6

)pp

471

-576

,20

12

If extra chromosomes reduce cell viability, how and why do aneuploid cancer cells thrive?

Centrosome amplification

Hyperstabilized kinetochore – microtubule interactions

C

D

As chromosomes are partitioned during cell division, errors can occur along several pathways, resulting in aneuploidy.

spindlecheckpointsystemallowcellstolive,butcauseaneuploidy,”saysYu.

Paradoxically,itturnsoutthattoomuchofaspindlecheckpointproteincanalsocauseaneuploidy.OverexpressionofMad2inmousecellscausesaneuploidy,afindingpublishedbyRocioSotillo,anHHMIinternationalearlycareerscientistattheEuropeanMolecularBiologyLaboratoryinMonterotondo,Italy,in2007.Theexcessproteininitiallyarrestscellsinthephaseofthecellcyclewhenthechromosomesarelinedupalongthecell’sequator.Afewcellsmanageto“escape”thatblock,saysSotillo,andproceedwithmitosis.Buttheyoverwhelminglyescapewithbaggage—extraorfewerchromosomes.

Othercelldivisiondefectsandspindlecheckpointerrorscauseaneuploidyaswell,andevidencefromorganismssuchasDrosophilaandplantsreaffirmsthattheresultsofaneuploidyareslowedcellproliferation,sickness,and

death.Unfortunately,theseobservationsstillfailedtoanswerthequestionthatnaggedatAmon,Pellman,andothers:ifextrachromosomesreducecellviability,howandwhydoaneuploidcancercellsthrive?

Elephant in the RoomIn1902,whileobservingtheabnormaldevelopmentofaneuploidcellsinseaurchinembryos,GermanzoologistTheodorBoveritheorizedthatatumormightdevelopfromsuchacell.Laterscientistsindeedfoundthatmosttumorsareaneuploid,butmanycastthefindingasideasanoddity.Fordecades,theroleofaneuploidyincancerwaslargelyignored.

“Ifyoulookatoldcancerreviews,theyneverputaneuploidyinthere,”saysStephenElledge,ageneticistandHHMIinvestigatoratHarvardMedicalSchool.“Ithasalwaysbeentheelephantintheroom.”Thatmaybebecause,duringthe1970sand1980s,itwassignificantlyhardertoidentifythecellulareffectsofawholeadditionalchromosome,asopposedtoasingleoncogene,Elledgesuggests.“Itjusthurtpeople’sheads.”

Withtheadventofnewimagingandsequencingtoolsbeginningintheearly1990s,aneuploidyreceivedrenewedattention—butsoonbecamestymiedbycontroversy.Leadingcancerresearcherssuggestedaneuploidywassimplyaby-productofcancer,anunimportantconsequence.Othersvehementlyarguedthatitcouldbethesolecauseofcancer.“Therewasa

18 Fall 2014 / HHMI Bulletin

hugedebate,”saysPellman.Butitdidn’tscarehimawayfromthefield.

Priortoworkingfulltimeinthelab,Pellmanwasapediatriconcologist.LikeAmon,aviewthroughamicroscopedrewhimtothestudyofaneuploidy:overandover,heobservedabnormalstructuresinthenucleusthatindicatedaneuploidyinthecellsofhisyoungcancerpatients.“Ispentalotoftimelookingattheseabnormalnuclearstructures,”saysPellman.“Therewasanobviousconnection.”

SohedecidedtoputoneaspectofBoveri’shypothesis—thatwholegenomeduplication,whichcauseshighratesofaneuploidy,mightcausecancer—tothetest.Pellmaninstructedoneofhispostdocs,TakeshiFujiwara,toproducetetraploid(twicethenormalchromosomenumber)mousecells,whichhaveahighrateofchromosomemis-segregationandthusrapidlybecomeaneuploid.Then,Fujiwarawastoinjectthosecellsintomice.AlthoughPellmanthoughtitwasimportanttotesttheidea,hefeltitwasalmosttoosimple,“toogoodto

betrue,”forthetetraploidcellstocausecancer.Fujiwara’sinitialresultsshowedthatthecellsindeedcausedaggressivetumorsinthemice.“Itwaswhatwewereafter,butitwasneverthelessverysurprisingthatitworkedsowell,”Pellmanrecalls.Thediscoveryprovidedstrongevidencethattetraploidy,andtheresultinganeuploidy,promotestumordevelopment.

Sincethen,Pellman’slabhasbeenworkingtodeterminehowaneuploidymightcausetumors.In2012,histeamshowedthatinsidemicronuclei—smallmembrane-boundbodiesthatpartitionextrachromosomesawayfromthecytoplasm—chromosomesaresubjecttoextensiveDNAdamage.Thissuggeststhatonewayaneuploidycausescanceristheold-fashionedway:throughmutationsthatactivatecancer-causinggenesorinactivatetumorsuppressors.Additionally,Pellman’slabrecentlydiscoveredevidenceshowingthatextracentrosomes,independentofgeneratinganeuploidy,canleadtotheactivationofRac1,aproteininvolvedincancercellsignalingandtumorinvasion(publishedJune5,2014,inNature).

Moreevidenceismountingthataneuploidypromotescancerbyalteringoncogenesandtumorsuppressors.LastyearinCell,ElledgeandcolleaguesatHarvardreportedthatcancergenomesselectforextrachromosomesthatcontainpotentoncogenes,andselectagainstextrachromosomesrichintumorsuppressors(see“TheMethodinCancer’sMadness,”Spring2014HHMI Bulletin).AttheMayoClinicinMinnesota,molecularbiologistJanvanDeursenandcolleagueshave

Aneuploidy—an abnormal number of chromosomes—appears to help a normal cell transform into a cancer cell. But not by giving it an edge to grow. Evidence suggests that aneuploidy actually slows cell division. So how exactly does aneuploidy benefit tumors?

Scientists have two main hypotheses. The first is that aneuploidy helps cancer cells adapt quickly under extreme pressure. In 2012, geneticist Yitzhak Pilpel and colleagues at the Weizmann Institute of Science in Israel tested this theory in yeast cells by exposing them to high heat. The yeast that acquired chromosome duplications—especially of chromosome III, which contains special proteins that aid in heat tolerance—survived better than

yeast that did not. When Pilpel reduced the temperature back to normal for these chromosome III aneuploids, the cells quickly lost the extra copy of the chromosome during subsequent replication.

Interestingly, however, when Pilpel kept the aneuploids at a high temperature, they also lost the extra copy, but only after evolving a different, more efficient way to express the heat-tolerance proteins over time. Pilpel and his team concluded that having an extra chromosome is a “quick fix,” but not a permanent solution, to deal with stress. That ability to quickly adapt would be a great benefit to cancer cells during metastasis. When a breast tumor cell spreads to the bone, for example, it must survive in a completely different

environmental niche. “As the disease progresses, adaptation potential provided by aneuploidy could be very beneficial,” says Angelika Amon, an HHMI investigator and yeast geneticist at MIT. The hypothesis has yet to be tested in cancer cells, however.

Another hypothesis is that aneuploidy promotes structural abnormalities known to cause cancer. In studies of yeast, Amon’s team has shown that aneuploidy causes genomic instability—a high frequency of mutations in the genome, from small mutations of two to three nucleotides to larger structural changes. Amon assayed 13 aneuploid yeast strains in a series of tests for different types of genomic instability, such as increased mutation rates and increased rates

of recombination. “It was really remarkable. Every aneuploid strain we looked at scored in at least one assay,” says Amon, and most tested positive for multiple types of genomic changes. The finding suggests that aneuploidy, no matter which chromosome is duplicated, causes genomic instability. And that instability predisposes the cell to mutations that cause cancer.

Aneuploidy may help cancer cells adapt quickly to stress in the body and nudge cells toward mutations that directly promote tumorigenesis, according to this recent work. “But so far, all this work is done in [cultured] cells, and that’s not good enough,” says Amon. Her team and others are planning to test these two hypotheses in mouse models of aneuploidy. — M.S.

Stressed-Out Cancer Cells: The Benefits of Aneuploidy

“We’re at the point where there’s pretty strong trust that aneuploidy is driving cancer.”—stephenelledge

19HHMI Bulletin / Fall 2014B

izu

ayeh

uT

esfa

ye/

AP

©H

HM

I

foundthataneuploidypredisposesmicewithonlyoneworkingcopyofp53,insteadofanormalpair,tolosethatcopyofthesuppressorgeneanddeveloptumors.

Basedoncurrentunderstanding,aneuploidyappearstopromotetumorproductionthroughmultiplemechanisms,butnotbyincreasingacell’sabilitytomultiply.That,atleast,helpsresolvetheaneuploidyparadox.“Wethink[aneuploidy]offerstumorcellssomethingotherthanproliferation,somethingelsethatisgoodfortumorigenesis,”saysAmon.(Seesidebar,“Stressed-OutCancerCells:TheBenefitsofAneuploidy.”)

Butdoesaneuploidycausecanceronitsown?VanDeursen’sworksuggeststhatitdoesnot.Formorethanadecade,vanDeursenhasbeenengineeringmicetoproducelowlevelsofindividualspindlecheckpointproteins,includingBub1andBubR1(completelossofsuchproteinsislethaltoembryos).Someofthemicedevelopcancer,fromlungcancertolymphomatolivertumors,butnotall.What’smore,vanDeursencan’tpredictwhichcheckpointproteinthatheknocksdown—andhehastriedadozen—willresultinatumor.

“Ifitweremerelyaneuploidythatdrivesmalignanttransformation,thenwe’dexpectthatallthemicewouldhavecancers,andmanycancers.Butthatreallyisn’tthecase,”saysvanDeursen.“It’slikelytheanswerisaneuploidyplussomethingelse.Soyouwillhavetwo,three,orfourindependent,tumor-promotingactivities.”

Today,mostresearchers,includingAmon,agreethatoncogenesarethemaindriversofcancer,andaneuploidycontributesvariationandinstabilitytothetumorcells.Ifoncogenesarethespark,aneuploidyisthekindling.“We’reatthepointwherethere’saprettystrongtrustthataneuploidyisdrivingcancer,”saysElledge.“Thequestionis,howpotentisit?”

Targeting StressUntilthatquestionisanswered,mostaneuploidyresearchremainsfocusedonbasicmechanisms.Butafewinquisitivescientistsaredippingtheirtoesintotranslationalresearch,lookingforwaystotargetorexploitaneuploidytopreferentiallykillcancercells.

“Aneuploidyisdefinitelystressingthecells,andthecellsmustdependonstresssupportpathwaysmorethannormalcellsdo,”saysElledge.“Soifyoucantinkerwiththat,youcanpotentiallymakethecellvulnerableorkillit.Thiscouldcertainlyaugmentothertherapies.”

SomeofPellman’sworksuggeststhattargetingcellswithtoomanycentrosomesmightbeawaytoselectivelytargetcancercells.SeveraldrugcompaniesarepursuingsmallmoleculedrugstoinhibitaspecificmotorproteinthatPellmanidentifiedasessentialforthesurvivalofextracentrosome-containingcancercells.

AtUTSouthwestern,Yuandcolleaguesrecentlycompletedanexperimentinwhichtheysilencedeachofthe20,000+genesinhumancancercells,onebyone,andthentreatedtheresultingcellswithTaxol.TheirgoalwastoseewhichgeneshelpcellsavoiddeathbyTaxol,whichisbelievedtocausemassiveaneuploidy,andwhichgenespromoteTaxol-triggereddeath.

Manyofthegenestheyidentifiedencodecomponentsofthespindlecheckpoint,suggestingpotentialdrugtargetstohelpTaxolworkbetter,saysYu.ThosetargetscouldleadtothekindofdrugsAmonhaslongdreamedof,toimprovetheeffectivenessofcurrentchemotherapyregimensandreducesideeffects.

BackatMIT,Amon’stranslationalworkisinfullswing.IncollaborationwithaHarvardscreeningfacility,Amonhasbeensearchingthroughlibrariesofcompoundsforchemicalsthatpreferentiallypreventaneuploidcellsfrommultiplying.Shehasidentifiedseveral.Oneofthemisachemicalinvolvedintheformationoflipidsthatappearstotargethighlyaneuploidcells.

Amon’steamfoundthatbycombiningTaxolwiththenewcompoundatconcentrationswhereeachcouldkill10percentofcancercellsifgivenseparately,themixturewasabletodestroycancercellswithanefficacyof90percent.“Nowwe’reworkingtotreatmousemodelsofcanceranddeterminethemechanism,”shesays.“We’rereallyexcitedaboutit.”

Stephen Elledge has found evidence that aneuploidy alters the balance of tumor suppressors and promoters.