potentialprojectace

DOI:10.
1126/science.
1188015,1492(2010);329Scienceetal.
MatthewG.
VanderHeiden,RapidlyProliferatingCellsEvidenceforanAlternativeGlycolyticPathwayinThiscopyisforyourpersonal,non-commercialuseonly.
.
clickingherecolleagues,clients,orcustomersby,youcanorderhigh-qualitycopiesforyourIfyouwishtodistributethisarticletoothers.
herefollowingtheguidelinescanbeobtainedbyPermissiontorepublishorrepurposearticlesorportionsofarticles(thisinformationiscurrentasofSeptember20,2010):Thefollowingresourcesrelatedtothisarticleareavailableonlineatwww.
sciencemag.
orghttp://www.
sciencemag.
org/cgi/content/full/329/5998/1492versionofthisarticleat:includinghigh-resolutionfigures,canbefoundintheonlineUpdatedinformationandservices,http://www.
sciencemag.
org/cgi/content/full/329/5998/1492/DC1canbefoundat:SupportingOnlineMaterialhttp://www.
sciencemag.
org/cgi/content/full/329/5998/1492#otherarticles,8ofwhichcanbeaccessedforfree:cites33articlesThisarticlehttp://www.
sciencemag.
org/cgi/collection/cell_biolCellBiology:subjectcollectionsThisarticleappearsinthefollowingregisteredtrademarkofAAAS.
isaScience2010bytheAmericanAssociationfortheAdvancementofScience;allrightsreserved.
ThetitleCopyrightAmericanAssociationfortheAdvancementofScience,1200NewYorkAvenueNW,Washington,DC20005.
(printISSN0036-8075;onlineISSN1095-9203)ispublishedweekly,exceptthelastweekinDecember,bytheScienceonSeptember20,2010www.
sciencemag.
orgDownloadedfromReferencesandNotes1.
V.
Ramanathan,P.
J.
Crutzen,J.
T.
Kiehl,D.
Rosenfeld,Science294,2119(2001).
2.
M.
L.
Ryder,J.
Langner,H.
Rodhe,Nature348,22(1990).
3.
M.
Chinetal.
,ClimateChangeScienceProgramSynthesisandAssessmentProduct2.
3,U.
S.
ClimateChangeScienceProgram,Ed.
(GlobalChangeResearchInformationOffice,Washington,DC,2009).
4.
S.
Twomey,R.
Gall,M.
Leuthold,Boundary-LayerMeteorol.
41,335(1987).
5.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
Atmos.
106,32555(2001).
6.
A.
Stohl,J.
Geophys.
Res.
Atmos.
111,1(2006).
7.
L.
Remeretal.
,J.
Geophys.
Res.
Atmos.
113,D14S07(2008).
8.
J.
E.
Penneretal.
,Atmos.
Chem.
Phys.
6,3391(2006).
9.
MeasurementsofCNcoldwereinitiallymadeonAAandPAusingaTSImodel3760(detectionlimit~12to14nm)andaTSImodel3010(detectionlimit~8to12nm)onotherexperiments.
Consequently,thelowvaluesofCNhotforAAandPAmaynotincludethesmallestnonvolatileresidualsizesbetween8and14nmdetectedonotherexperiments.
CNhotparticlestypicallyincludesoot,dust,seasalt,and/orsomeorganicaerosolstableat300°C.
Residencetimesintheheaterforexperimentsweretypically1to2s,and,evenif99%ofthemassisvolatilized,somelargersizes>100-nmdiametermaynotfullyvaporize.
A100-nmparticlewithonly0.
1%massremainingat350°Cwillstillbe10nm.
However,mostsmallervolatilesizeswillshrinkbelowthedetectionlimitsuchthatCNhotcommonlyselectsforlargersizeslikelytobeCCN.
10.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
Atmos.
104,5735(1999).
11.
A.
D.
Clarke,V.
N.
Kapustin,J.
Atmos.
Sci.
59,363(2002).
12.
F.
Raes,J.
Geophys.
Res.
Atmos.
100,2893(1995).
13.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
Atmos.
112,D12S18(2007).
14.
J.
G.
Hudson,Geophys.
Res.
Lett.
34,L08801(2007).
15.
W.
A.
Hoppel,G.
M.
Frick,J.
W.
Fitzgerald,R.
E.
Larson,J.
Geophys.
Res.
Atmos.
99,14443(1994).
16.
V.
N.
Kapustinetal.
,J.
Geophys.
Res.
Oceans111,1(2006).
17.
T.
L.
Andersonetal.
,J.
Geophys.
Res.
108,ACE15(2003).
18.
J.
E.
Penneretal.
,J.
Atmos.
Sci.
59,441(2002).
19.
A.
C.
Staudtetal.
,J.
Geophys.
Res.
Atmos.
106,32571(2001).
20.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
Atmos.
103,16397(1998).
21.
Y.
Shinozuka,A.
D.
Clarke,S.
G.
Howell,V.
N.
Kapustin,B.
J.
Huebert,J.
Geophys.
Res.
Atmos.
109,D24201(2004).
22.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
103,16397(1998).
23.
J.
G.
Hudson,Y.
Xie,S.
S.
Yum,J.
Geophys.
Res.
Atmos.
103,16609(1998).
24.
E.
V.
Browelletal.
,J.
Geophys.
Res.
Atmos.
106,32481(2001).
25.
K.
G.
Moore,A.
D.
Clarke,V.
N.
Kapustin,S.
G.
Howell,J.
Geophys.
Res.
108,8236(2003).
26.
D.
J.
Jacobetal.
,J.
Geophys.
Res.
108,GTE2-1(2003).
27.
B.
J.
Huebertetal.
,J.
Geophys.
Res.
108,ACE1-1(2003).
28.
Y.
Maetal.
,J.
Geophys.
Res.
108,GTE37-1(2003).
29.
A.
D.
Clarkeetal.
,J.
Geophys.
Res.
109,D15S09(2004).
30.
M.
Chinetal.
,J.
Geophys.
Res.
108,ACE22-1(2003).
31.
Y.
Shinozukaetal.
,J.
Geophys.
Res.
Atmos.
112,D12S20(2007).
32.
C.
M.
Carrico,P.
Kus,M.
J.
Rood,P.
K.
Quinn,T.
S.
Bates,J.
Geophys.
Res.
108,ACE18-1(2003).
33.
R.
Halthore,P.
F.
Caffrey,J.
Geophys.
Res.
33,L14819(2006).
34.
C.
S.
McNaughtonetal.
,J.
Geophys.
Res.
Atmos.
109,D19S06(2004).
35.
L.
Jaegleetal.
,J.
Geophys.
Res.
108,GTE23-1(2003).
36.
C.
L.
Healdetal.
,J.
Geophys.
Res.
108,4804(2003).
37.
A.
D.
Clarke,K.
J.
Noone,Atmos.
Environ.
19,2045(1985).
38.
M.
O.
Andreae,Atmos.
Chem.
Phys.
9,543(2009).
39.
D.
A.
Hegg,H.
Jonsson,J.
Geophys.
Res.
Atmos.
105,1987(2000).
40.
TheSOMincludesaltitude-resolvedmeandataforkeyparametersrelatedtothispaper.
Additionalhigher-resolutiondataformostexperimentsdiscussedherecanbefoundatthefollowinglinks:forPEMT-A,PEMT-B,andTRACE-P,http://www-gte.
larc.
nasa.
gov/FM.
html;forACE-Asia,http://data.
eol.
ucar.
edu/master_list/project=ACE-ASIA;forINTEX-A,http://www-air.
larc.
nasa.
gov/cgi-bin/arcstat;forINTEX-B(IBM,IBH,IBA),IMPEX,andMIRAGE,http://www-air.
larc.
nasa.
gov/cgi-bin/arcstat-b;andforARCTAS,http://www-air.
larc.
nasa.
gov/cgi-bin/arcstat-c.
41.
WethanktheNSFAtmosphericChemistryProgram,theNASAGlobalTroposphericExperiment,andNASAEarthScienceProjectOfficeforthemanyyearsofsupportrepresentedhere.
WealsothankG.
Sachse,G.
Kok,T.
Campos,andJ.
PodolskefortheirassociatedCOmeasurementsandourcolleaguesS.
Howell,C.
McNaughton,Y.
Shinozuka,J.
Zhou,J.
Porter,B.
Lieneart,M.
Litchy,K.
Moore,andM.
Pinkertonfortheirdedicatedeffortsinobtainingthesedata.
ThisretrospectivesynthesisissupportedundergrantsNASA657001and652485andNSFATM05-11521.
ThisisSOESTcontributionnumber7948.
SupportingOnlineMaterialwww.
sciencemag.
org/cgi/content/full/329/5998/1488/DC1TablesS1toS1426February2010;accepted29July201010.
1126/science.
1188838EvidenceforanAlternativeGlycolyticPathwayinRapidlyProliferatingCellsMatthewG.
VanderHeiden,1,2,3*JasonW.
Locasale,2,3KennethD.
Swanson,2HadarSharfi,2GregJ.
Heffron,4DanielAmador-Noguez,5HeatherR.
Christofk,2GerhardWagner,4JoshuaD.
Rabinowitz,5JohnM.
Asara,2LewisC.
Cantley2,3Proliferatingcells,includingcancercells,requirealteredmetabolismtoefficientlyincorporatenutrientssuchasglucoseintobiomass.
TheM2isoformofpyruvatekinase(PKM2)promotesthemetabolismofglucosebyaerobicglycolysisandcontributestoanabolicmetabolism.
Paradoxically,decreasedpyruvatekinaseenzymeactivityaccompaniestheexpressionofPKM2inrapidlydividingcancercellsandtissues.
Wedemonstratethatphosphoenolpyruvate(PEP),thesubstrateforpyruvatekinaseincells,canactasaphosphatedonorinmammaliancellsbecausePEPparticipatesinthephosphorylationoftheglycolyticenzymephosphoglyceratemutase(PGAM1)inPKM2-expressingcells.
WeusedmassspectrometrytoshowthatthephosphatefromPEPistransferredtothecatalytichistidine(His11)onhumanPGAM1.
ThisreactionoccurredatphysiologicalconcentrationsofPEPandproducedpyruvateintheabsenceofPKM2activity.
Thepresenceofhistidine-phosphorylatedPGAM1correlatedwiththeexpressionofPKM2incancercelllinesandtumortissues.
Thus,decreasedpyruvatekinaseactivityinPKM2-expressingcellsallowsPEP-dependenthistidinephosphorylationofPGAM1andmayprovideanalternateglycolyticpathwaythatdecouplesadenosinetriphosphateproductionfromPEP-mediatedphosphotransfer,allowingforthehighrateofglycolysistosupporttheanabolicmetabolismobservedinmanyproliferatingcells.
Oneofthemajordifferencesobservedbe-tweencancercellsandnormalcellsisinhowtheymetabolizeglucose;mostcancercellsprimarilymetabolizeglucosebyglycolysis,whereasmostnormalcellscompletelycatabolizeglucosebyoxidativephosphorylation(1).
Thisshifttoaerobicglycolysiswithlactateproduction(alsoknownastheWarburgeffect),coupledwithincreasedglucoseuptake,islikelyusedbypro-liferatingcellstopromotetheefficientconver-sionofglucoseintothemacromoleculesneededtoconstructanewcell(2).
TheglycolyticenzymepyruvatekinaseisalternativelysplicedtoproduceeithertheM1(PKM1)orM2(PKM2)isoforms(3).
Thespliceisoformofpyruvatekinaseex-pressedincellsinfluencestheextenttowhichglucoseismetabolizedbyeitheraerobicglycolysisoroxidativephosphorylation.
CellsexpressingPKM2producemorelactateandconsumelessoxygenthancellsexpressingPKM1(4).
Consist-entwiththismetabolicphenotype,allcancercellsstudiedtodateexclusivelyexpressPKM2,where-ascellsinmanynormaldifferentiatedtissuesex-pressPKM1.
PKM2differsfromPKM1inthatitsactivitycanbenegativelyregulatedinresponsetogrowthfactorsignalingbybindingtotyrosine-phosphorylatedproteins(5,6).
Paradoxically,itisthisabilitytointeractwithtyrosine-phosphorylatedproteins,andtodecreasepyruvatekinaseactivity,thatappearstobeimportantforcellproliferation(5).
Thisselectionforthedecreasedactivityofarate-limitingglycolyticenzymeappearsinconsistentwiththeincreasedglucoseusethatischarac-teristicofcancercells.
However,completecatabo-lismofpyruvatetoCO2maybecounterproductiveinadividingcellbecauseitmaylimittheavailability1DanaFarberCancerInstitute,HarvardMedicalSchool,Boston,MA02115,USA.
2BethIsraelDeaconessMedicalCenter,DivisionofSignalTransductionandDepartmentofMedicine,HarvardMedicalSchool,Boston,MA02115,USA.
3DepartmentofSystemsBiology,HarvardMedicalSchool,Boston,MA02115,USA.
4DepartmentofBiologicalChemistryandMolecularPhar-macology;HarvardMedicalSchool,Boston,MA02115,USA.
5Lewis-SiglerInstituteforIntegrativeGenomicsandDepartmentofChemistry,PrincetonUniversity,Princeton,NJ08544,USA.
*Presentaddress:KochInstituteforIntegrativeCancerResearchatMassachusettsInstituteofTechnology,Cambridge,MA02139,USA.
Towhomcorrespondenceshouldbeaddressed.
E-mail:lewis_cantley@hms.
harvard.
edu17SEPTEMBER2010VOL329SCIENCEwww.
sciencemag.
org1492RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromofprecursorsandreducingpotentialnecessarytoproducebiomass.
PKM2islessactivethanPKM1invitroandincells.
WedirectlycomparedthespecificactivityofPKM1andPKM2bothinvitroandincelllysates(7).
RecombinantPKM1enzymehadahighspecificactivitythatwasindependentofthePKM2-specificallostericactivatorFBP(Fig.
1A)(8).
ThespecificactivityofPKM2thatisfullyactivatedbyFBPisabouthalfthatofPKM1.
ThepropertyofPKM2thatappearstopromotecellproliferationinvivoisitsinterac-tionwithtyrosine-phosphorylatedproteinsandconsequentreleaseofFBP.
IntheabsenceofFBP,PKM2hadlessthanone-quarteroftheactivityofPKM1(Fig.
1A).
Todeterminewhetherthedif-ferencesinactivityobservedwithrecombinantenzymesarealsoseenincells,wemeasuredpyru-vatekinaseactivityinlysatesfromcellsengi-neeredtoexpressequivalentamountsofeitherPKM1orPKM2intheabsenceoftheotherisoform(Fig.
1B).
Undertheseidenticalconditions,PKM2expressionprovidesaselectiveadvantageforgrowthinvivo(4).
LysatesfromPKM2-expressingcellsexhibitedlessthanhalfthepyruvatekinaseactivityoflysatesfromcellsexpressingtheequiv-alentamountofPKM1(Fig.
1C).
Thus,theselec-tionforPKM2expressioninproliferatingcellsisaccompaniedbyadecreaseintotalcellularpyruvatekinaseactivity.
Donationofphosphatefromphosphoenol-pyruvatetoacytosolicproteinofabout25-kD.
ItispossiblethattherelativedecreaseinPKM2activityallowsanupstreammetaboliteinglycol-ysistosignalenergystatusortobeshuntedtoanundiscovered,orunderappreciated,metabolicpathwayrequiredforcelldivision.
Thesubstrateforpyruvatekinaseincellsisphosphoenolpyr-uvate(PEP).
BacteriausePEPastheinitialphos-phatedonorforproteinphosphorylationinasignalingcascadethatregulatescarbohydratemetabolisminresponsetonutrientavailability(9,10).
Inaddition,transferofthePEPphosphatetoaproteinoccursasanenzymaticintermediatewithintheCalvincycleofC4plants(11,12).
ThispromptedustoexplorethepossibilitythatPEPmighttransferitsphosphatetoaproteininmam-maliancells.
Wegenerated[32P]PEP(fig.
S1)andtestedhypotoniclysatesfromhumanembry-onickidney(HEK)cellsforthepresenceofsuchaPEP-dependentproteinphosphorylationactivity.
Incubationofextractswith[g-32P]adenosinetriphosphate(ATP)resultedinnumerous32P-labeledproteins,andthe32P-labelingoftheseproteinswasabolishedafteradditionofa100-foldexcessamountofnonradioactiveATP(Fig.
1D).
Nodecreaseinincorporationofphosphatefrom[g-32P]ATPwasobservedinthepresenceofexcessnonradioactivePEP.
However,incubationofcellextractswith[32P]PEPresultedintheincorporationof32Pintoseveralproteins,themostprominentofwhichresolvedatarelativemolecularsizeofabout25-kDbySDS–polyacrylamidegelelectrophoresis(PAGE).
The32P-labelingofthisproteinwaseliminatedafteradditionofexcessamountsofnonradio-activePEPbutnotbyexcessnonradioactiveATP,consistentwithPEPactingasthephosphatedonor(Fig.
1D).
Otherpurinenucleotides,includingguaninetriphosphate(GTP),didnotcompetewith32P-labeledPEPtophosphorylatethe25-kDprotein,andphosphorylationofa25-kDproteinwasobservedinextractsfrommultiplecelllinesincubatedwith[32P]PEP(fig.
S2).
CytoplasmicconcentrationsofPEParelessthan30mMineukaryoticcells(13,14).
TodeterminewhetherFig.
1.
EvidenceofPEP-dependentphosphoryl-ationofa25-kDproteininPKM2-expressingcellswithlesspyruvatekinaseactivity.
(A)6*His-taggedhumanPKM1andPKM2wereexpressedinEsche-richiacoliandpurifiedbyNiaffinitychromatog-raphy.
Thespecificac-tivityofeachenzymewasdeterminedinthepresenceofsaturatingamountsofPEPandadenosinediphosphate(ADP).
TheactivityofPKM1andPKM2inthepresenceandabsenceofFBPisshown.
ErrorbarsindicateSEMinallfig-ures.
(B)H1299cellswereengineeredtoexpressequiv-alentamountofPKM1orPKM2proteinasdescribedpreviously(4).
EquivalentexpressionofPKM1andPKM2wasconfirmedbyWesternblotusinganantibody(aPK)thatrec-ognizesanepitopesharedbyPKM1andPKM2.
(C)Asin(A),pyruvatekinaseactivitywasdeterminedbyusingsaturatingamountsofPEPandADP.
TherelativepyruvatekinaseactivityobservedinthePKM1-orPKM2-expressingcellsdescribedin(B),relativetolysisbufferalone,isshown.
(D)HEK293cellswerehypotonicallylysedandincubatedwith[32P]ATPor[32P]PEPbeforeanalysisbySDS-PAGEandauto-radiography.
Thelysateswereincubatedwith[32P]ATPor[32P]PEPinthepresenceof10mMATPorPEP,respectively(–),orwiththeadditionof1mMnonradio-activecompetitorATPorPEP.
(E)Celllysatewasincubatedwith[32P]PEPinthepresenceoftheindicatedconcentrationofnonradioactivecompetitorPEPbeforeanalysisbySDS-PAGEandautoradiography.
(F)Celllysatewasincu-batedwith32P-labeledPEPasabove,andthepHofthereactionwaschangedtopH1orpH13.
Reactionswereincubatedfor2hoursat65°CbeforeanalysisbySDS-PAGEandautoradiography.
01002003004005006007008009001000M1M2Specificactivity(nmol/s/mgenzyme)-FBP+FBPA020040060080010001200140016001800NoLysatemPKM1mPKM2RelativePKactivityC-ATPPEP-ATPPEPExcesscompetitor32P-ATP32P-PEPDCtrl(10M)100050025012562.
531.
3evitcaoidar-noNcompetitorPEP(M)ECtrlpH1pH1332P-PEPFBmPKM1mPKM2PKwww.
sciencemag.
orgSCIENCEVOL32917SEPTEMBER20101493RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromphosphorylationofthisproteinispossibleatlowmMconcentrationsofPEP,weaddedin-creasingconcentrationsofnonradioactivePEPtoestimatetheMichaelisconstant(Km)forPEPinthereaction(Fig.
1E).
IncreasingtheamountofunlabeledPEPabove10mMresultedinde-creased32P-labelingofthe25-kDprotein,suggest-ingthattheKmforPEPinvolvedinthisreactionisinarangewherethisreactioncouldoccuratconcentrationsofPEPpresentincells.
PEP-dependentphosphorylationofthe25-kDproteinonhistidine.
Thephosphorylationreac-tioninvolvingPEPinbacterialtwo-componentsignalingandtheanalogousPEP-dependentproteinphosphorylationasanenzymaticinter-mediateinC4plantsbothinvolvetransferofthePEPphosphatetoahistidineresidue.
O-linkedproteinphosphatessuchasthoseseenuponphos-phorylationofserine,threonine,ortyrosineresiduesarestableunderacidicconditions(15,16).
Incon-trast,N-linkedproteinphosphatessuchasphos-pholysineorphosphohistidinearelabileatlowpHbutstableunderbasicconditions.
Therefore,weexposedproteinextractsthathadbeenincubatedwith[32P]PEPtoacidicorbasicconditionsbeforeanalysisbySDS-PAGEandautoradiography.
The32PsignalwaslostafterincubationatpH1butretainedatpH13(Fig.
1F),whichisconsistentwithanN-linkedphosphateresultingfromPEP-dependentphosphorylationofthe25-kDprotein.
Incubationof[32P]ATP-labeledlysatesatpH1re-sultedinnolossofproteinphosphorylation,indicat-ingthatlossofsignalfromthePEP-phosphorylatedproteinatlowpHwasnottheresultofnonspecificacidhydrolysis.
ConsistentwithaN-linkedphos-phorylation,in"standard"phosphoaminoacidanalysisinvolvingacidhydrolysisofthe25-kDPEP-phosphorylatedproteinalloftheresultingradioactivitymigratedasinorganicphosphate(Pi)onthin-layerelectrophoresis(fig.
S3A).
Duringreverse-phasethin-layerchromatographyafterbasehydrolysis,the32Pmigratedwithphospho-histidine,consistentwithhistidineasthetargetofPEP-dependentphosphotransfer(fig.
S3B).
Identificationofthe25-kDPEP-phosphorylatedproteinasphosphoglyceratemutase.
ThePEP-utilizingphosphorylatingactivityandthe25-kDtargetofphosphorylationwerepresentintheFig.
2.
PGAM1asthetargetofPEP-dependentphosphorylationthroughanenolase-independentreaction.
(A)TheS100fractionfromaHEK293celllysateswaspassedsequentiallythroughacustomcolumnandastrongcationexchangecolumnbeforeincubat-ingwith[32P]PEP(SFT).
Thisreactionwasthenappliedtoahydroxy-apatite(HAP)columnandelutedwith50mMNaHPO4.
Thesaltelution(E)containingthe32P-labeledspecieswasdilutedto<25mMNaHPO4andappliedtoaweakanionexchange(DEAE)column.
ElutionfromtheDEAEcolumnwasperformedwith100mMand200mMNaCl.
The200-mMsaltfractioncontainingthe32P-labeledspecieswasdilutedto50mMNaClandappliedtoastronganionexchange(Q)columnandelutedsequentiallywith100mMand350mMNaCl.
The350-mMsaltfractioncon-tainingthe32P-labeledspecieswasacetone-precipitatedforanaly-sisby2D-IEFandSDS-PAGE.
AnaliquotofeachfractionwasanalyzedbySDS-PAGEandautoradiography.
Flow-throughfractionsareindicatedasFT.
(B)Theacetone-precipitated350mMsaltfractiondescribedin(A)wasseparatedby2D-IEFandSDS-PAGE,andthe32P-labeledspecieswasidentifiedbyautoradi-ography.
(C)Theacetone-precipitated350-mMsaltfractionpreparedasdescribedin(A)wasseparatedby2D-IEFandSDS-PAGE,andproteinswereidentifiedbyCoomassiestain.
Thespeciescorrespondingtothe32P-labeledspeciesisindicatedwithanarrow.
(D)HEK293cellsweretransientlytransfectedwithcontrolplasmid(Control),aN-terminallyFLAG-taggedPGAM1comple-mentaryDNA(cDNA)(N-FLAGPGAM1),oraC-terminaltripleFLAG-taggedPGAM1cDNA(3*C–FLAGPGAM1).
Hypotoniclysatesfromthesecellswereincubatedwith[32P]PEPalone(Ctrl)orinthepresenceof1mMcoldcom-petitorATPorPEP.
TheproductsofthesereactionswereseparatedbySDS-PAGEandanalyzedbyautoradiography.
ProteinimmunoprecipitatedwithanantibodytoFLAGfromthereactionswithoutnonradioactivecompetitorwerealsoanalyzedbySDS-PAGEandautoradiography.
(E)Recombinant6*His-taggedPGAM1(rPGAM1)wasproducedinE.
coliandpurifiedbyNiaffinitychromatography.
IncreasingquantitiesofrPGAM1wereincubatedwith10mgofHEK293celllysateand[32P]PEP.
Thephosphorylationofboththeendog-enousPGAM1presentinthecelllysateandrPGAM1wasdeterminedbySDS-PAGEandautoradiography.
(F)Celllysateswereincubatedwith[32P]PEPintheabsence(Ctrl)orpresenceofNaForexogenouslyaddedrabbitmuscleenolaseenzyme(Eno).
ThelabelingofPGAM1wasdeterminedbySDS-PAGEandautoradiography.
75SDSPAGEIEF+-50353025SDSPAGE50353025IEF+-SFTFTEHAPDEAEQAcetoneprecipitateFT100200FT100350Ctrl+ATP+PEPCtrl+ATP+PEPCtrl+ATP+PEPcompetitorcompetitorcompetitorCtrlN-FLAG3xC-FLAGFLAGIPControlN-FLAGPGAM13xC-FLAGPGAM1ACDBCtrl+NaF+EnoPGAM1ErPGAM(ng)rPGAM1EndogenousCtrl18.
5551665001500F17SEPTEMBER2010VOL329SCIENCEwww.
sciencemag.
org1494RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromS100cytosolicfractionofHEK293cellsandwereretainedintheflow-throughfractionsofbothastrongcationexchangecolumnandaC11NPcolumn,whichcontainedaresinwesyn-thesizedasapossiblePEPaffinitycolumn.
Wefractionatedthepreviouslyphosphorylated25-kDtargetbybothanionexchangeandhydroxy-apatitechromatographyusingsaltelutionbutwereunabletodetectphosphorylationactivitywithinanyofthefractionsfromthesecolumnsupontheirincubationwith[32P]PEP.
Thisiscon-sistentwiththepresenceofanenzymatictransferofphosphatefromPEPtoaproteinandasep-arationofthe25-kDtargetproteinfromanactivityrequiredfortheobservedphosphoryl-ation.
WepurifiedthetargetproteinbycollectingtheflowthroughfractionsfromanS100proteinextractpassedsequentiallythroughaC11NPcol-umnandastrongcationexchange(S)column.
The[32P]PEPwasaddedtothispartiallyfrac-tionatedlysatetolabelthe25-kDprotein,andtheresulting32P-labeledproteinwasthensequen-tiallyfractionatedoverthreecolumnstoeffectthemaximalrecoveryofthetargetprotein(Fig.
2A).
Theproteinsinthefractionfromthestronganionexchange(Q)columncontainingthe32P-labeled25-kDproteinwereprecipitatedwithacetoneandsubjectedtotwo-dimensional(2D)electrophore-sis(Fig.
2,AandB,andfig.
S4).
Analysisofthe2Dsequentialisoelectricfocusing(IEF)andSDS-PAGEgelbyautoradiographyidentifiedasingleradioactiveisoelectricspecies(pI=~6.
2)at25-kD(Fig.
2B).
The32P-labeledspotcorrespondedtothemostacidicCoomassie-stainedspotinaseriesofseveraladjacentisoelectricspeciesat25-kDonaCoomassie-stainedgel(Fig.
2C).
In-geltryp-sindigestionfollowedbymicrocapillaryliquidchromatography–tandemmassspectrometry(LC/MS/MS)withproteindatabasesearchingidenti-fiedthe32P-labeledspeciesaswellastheadja-centspeciestobedifferentisoelectricformsoftheglycolyticenzymephosphoglyceratemutase1(PGAM1)(tableS1).
ToconfirmthatPGAM1wasindeedthetargetofPEPphosphorylation,wetransientlytransfectedFLAG-taggedPGAM1constructsintoHEK293cellsandincubatedlysatesfromthesecellswith[32P]PEP.
WhentheproteinsinthesereactionswereanalyzedbySDS-PAGEandautoradi-ography,asecond32P-labeledspeciesofgreatermolecularweightcorrespondingtothesizeoftheepitope-taggedPGAM1wasobserved(Fig.
2Dandfig.
S5A).
ThelargerspecieswasremovedandrecoveredbyimmunoprecipitationwithanantibodytoFLAG,anditslabelingwith32PwasblockedwithexcessnonradioactivePEPbutnotwithexcessnonradioactiveATP(Fig.
2D).
Thus,the32Pfrom[32P]PEPcanbetransferredtoPGAM1.
Toconfirmthatthe25-kDproteinlabeledfrom[32P]PEPisalsoPGAM1,weincubatedlysatesfromcontrol-andepitope-taggedPGAM1-transfectedcellswith[32P]PEPandsubjectedthemtolimitedproteolysis.
WhenanalyzedbySDS-PAGE,boththecontrollysatesandthosecontainingFLAG-taggedPGAM1producedidenticalpatternsof32P-labeledpeptidesafterlimitedproteolysis(fig.
S5B).
Lastly,recombinantPGAM1addedwith[32P]PEPtoafixedamountofcelllysatecouldcompeteforphosphorylationofendogenousPGAM1(Fig.
2E).
ThesedatademonstratethatPGAM1canbephosphorylatedbyPEP.
PGAM1phosphorylationisindependentofenolaseactivity.
AlthoughrecombinantPGAM1canbephosphorylatedby[32P]PEPincelllysates,purifiedrecombinantPGAM1isnotadirectsub-strate(fig.
S5C),indicatingthatanenzymaticac-tivityincelllysatesisrequiredtocatalyzePGAM1phosphorylationbyPEP.
PGAM1actsasanen-zymeinglycolysistointerconvert3-phosphoglycerate(3PG)and2-phosphoglycerate(2PG)throughaphosphohistidineintermediate(17).
BecausePEPcanbeconvertedbytheglycolyticenzymeeno-laseinto2PG,itseemedpossiblethatthePEP-dependentphosphorylationofPGAM1thatweobservedincelllysatesinvolvedconversionof[32P]PEPinto[32P]2PGbyenolase,followedbytransferofthe[32P]phosphatefrom2PGtothecatalytichistidineofPGAM1.
Wethereforein-creasedenolaseactivitybyadditionofexogenousenolaseenzymeordecreaseditbyadditionoftheenolaseinhibitorNaF(18)tocelllysates(fig.
S6).
TheinhibitionofenolaseactivitywithNaFhadminimaleffectonthetransferof32PfromPEPtoPGAM1(Fig.
2F).
Furthermore,theadditionofexogenousenolasepreventedthetransferof32PfromPEPtoPGAM1presumablybyconverting[32P]PEPto2PG.
Thesedataindicatethatconver-sionofPEPto2PGbyenolaseisnotinvolvedintheobservedphosphotransferfromPEPtoPGAM1.
PhosphorylationofPGAM1onthecatalytichistidine(His11)bythephosphatefromPEP.
TodeterminewhetherthephosphatefromPEPwastransferredtooneormoresitesonPGAM1,welabeledPGAM1with32PfromPEP,digestedtheproteinwithtrypsin,andanalyzedtheresultingpeptidesby2Dthin-layerchromatographyandthin-layerelectrophoresisfollowedbyautoradi-ography(fig.
S7A).
Thisrevealedasingle32P-labeledspecies,indicatingthatonlyasinglesiteisphosphorylatedonPGAM1inthereactionwithPEP.
PhosphoaminoacidanalysisindicatedthatPGAM1wasphosphorylatedonahistidineresi-due(fig.
S3).
Todeterminewhichhistidineresiduewasphosphorylated,weincubatedrecombinantPGAM1with[32P]PEPandHEK293celllysateandrecoveredthe[32P]PGAM1throughasso-ciationwithNi-agarosebeads(fig.
S7B).
This[32P]PGAM1wasthendigestedwithtrypsin,andtheresultingpeptideswereseparatedbyhigh-performanceliquidchromatography(HPLC)(Fig.
3A).
Asingle32P-labeledpeptidewasobserved.
HPLCfractionswerecollectedtoconfirmwhichpeakcontainedthe32P-labeledpeptide(fig.
S7,CandD),andthefractioncontainingthe32P-labeledpeptidewassequencedbyLC/MS/MSusingahybridlineariontrap–orbitrapmassspectrometerwithuseofthehigherenergycollisiondissociationcell(HCD)(19).
HCDwasrequiredtoclearlyresolvethelowmassfragmentionstoshowthatthesiteofphosphorylationinthepeptidesequencewaslocalizedtoHis11(H11)ofPGAM1(Fig.
3Bandfig.
S8).
Detectionofhistidinephosphoryl-ationusingmassspectrometryischallengingbuthaspreviouslybeenreported(20–23).
ThepHissitewasconfirmedbyusingtwodifferentcom-merciallyavailabledatabasesearchalgorithms[Mascot(24)andSequest(25)]withstatisticallysignificantscores.
ConsistentwithH11beingtheresiduephosphorylated,mutationofH11toAsp11abolishedtransferof32PfromPEPtoPGAM1(Fig.
3C).
ToconfirmthatthephosphateatH11isfromexogenouslyaddedPEPratherthanaphosphatethatwaspresentbeforecelllysis,weincubatedrecombinantPGAM1with[18O]phosphate-labeledPEP(fig.
S9)inthepresenceof1mMnormalisotopicATPandHEK293celllysateandthenisolatedPGAM1withNi-agarosebeads,digesteditwithtrypsin,andseparatedthepeptidesbyHPLC.
ThepeptidefractioncontainingH11wasanalyzedbyorbitrapmassspectrometryinFouriertransformmassspectrometry(FT-MS)mode,andseveralisotopicspecieswereidentifiedthatcor-respondedtotheH11-phosphorylatedpeptide(Fig.
3D).
Theheavyisotopicformswereconsistentwith18Olabelingofthephosphatethatwastrans-ferredtothepeptidefromthe[18O]PEPratherthanfromthenormalisotopic[16O]ATP.
ThesedatademonstratethatthephosphategroupfromPEPistransferredtoH11ofPGAM1.
AssociationofPGAM1phosphorylationwithpyruvategenerationfromPEPintheabsenceofpyruvatekinase.
Becausethe[32P]phosphate(and[18O]phosphate)fromPEPistransferredtothecatalytichistidineofPGAM1,wewonderedwhetherwewereobservinganetincreaseinH11phosphorylationofPGAM1ormerelyobservinganexchangeofphosphatealreadypresentonPGAM1withphosphatefromPEP(ascanoccurduringtheinterconversionof2PGand3PG).
Toaddressthisissue,weaddedrecombinantPGAM1toacellextractinthepresenceorabsenceofPEP.
Theseextractswerethensubjectedto2DgelelectrophoresisandanalyzedbyWesternblotwithanantibodytoPGAM.
ConsistentwithPEPphosphorylationofH11,anew,moreacidiciso-electricspeciesofbothendogenousandrecombi-nantPGAM1wasdetectedinthePEP-containinglysate(Fig.
4A).
NochangeintheisoelectricformsofPGAM1wasobservedwhenlysateswereincu-batedwithATPinsteadofPEP(fig.
S10A).
ToconfirmthatthisspeciesdidindeedrepresenttheH11-phosphorylatedformofPGAM1,weaddedPEPtoacelllysatetophosphorylatePGAM1andthenincubatedthereactioneitheratneutralpHoratpH2tochemicallydisruptH11phos-phorylation.
Analysisof2DWesternblotswithanantibodytoPGAM1showedthatincubationatpH2resultedinlossofthemostacidicPGAM1species(fig.
S10B).
Therefore,weconcludedthat2DWesternblotscouldbeusedtoassessH11phosphorylationstatusincells.
ThesedataalsodemonstratethatPEPcancauseanetincreaseinH11-phosphorylatedPGAM1andthatthephos-phorylationofPGAMweobservecannotbeac-www.
sciencemag.
orgSCIENCEVOL32917SEPTEMBER20101495RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfrom17SEPTEMBER2010VOL329SCIENCEwww.
sciencemag.
org1496RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromcountedforbyexchangeofthePEPphosphatewithapreviouslyphosphorylatedH11.
TodeterminewhetherH11-phosphorylatedPGAM1iscatalyticallycompetentforenzymaticactivity,weassayedtheabilityofPEP-phosphorylatedPGAM1toconvert3PGto2,3-bisphosphoglycerate(2,3-BPG),theintermediatein3PGto2PGcon-version(26).
RecombinantHis-taggedPGAM1wasincubatedwithPEPandcellextracttoallowphosphorylationonH11,andtheproteinwasre-coveredthroughassociationwithNi-agarosebeads.
Additionof3PGtotherecoveredPGAM1resultedin2,3-BPGproductionasdeterminedbyselectedreactionmonitoring(SRM)usinghybridquadrupolelineariontrapmassspectrometry(fig.
S11).
Thus,phosphorylationofPGAM1byPEPleadstoanenzymespeciesthatisactivetocarryouttheknownenzymaticfunctionofPGAM1.
WefractionatedacelllysateoveraweakanionexchangecolumnandisolatedthePEP-dependentPGAM1-phosphorylatingactivityinafractionthatwasseparatefromtheenolase-containingfractionasdeterminedbybothen-zymeactivityassaysandWesternblot(Fig.
4,BandC,andfig.
S12A).
ThefractioncontainingthePGAM1-phosphorylatingactivitywasalsoseparatedfrompyruvatekinase,asdeterminedbybothenzymeactivityassayandWesternblot.
InsupportofthisfindingthatpyruvatekinaseisnotinvolvedinthetransferofphosphatetoPGAM1,wefoundthatshorthairpinRNAknockdownofpyruvatekinaseresultedintheenhancedabilityofacelllysatetotransfer32Pfrom32P-labeledPEPtoPGAM1withnochangeinthelevelofPGAM1protein(fig.
S12,BtoD).
Ithasbeenreportedthatacomplexcontainingthenucleosidediphosphatekinasenm23andglyceraldehyde-3-phosphatedehydrogenase(GAPDH)canphospho-rylatePGAM1(27).
However,neitherGAPDHnornm23copurifyinsubstantialquantitieswiththePEP-dependentPGAM1-phosphorylatingac-tivity(fig.
S12E),suggestingthattheseproteinsarenotinvolvedintheactivityweobserve.
WefurtherinvestigatedtheconsequencesofmetabolizingPEPthroughphosphotransfertoPGAM1.
TotestwhetherPEPisconvertedtopyruvateduringthephosphotransferreaction,weincubatedtheanionexchangefractioncontain-ingthePGAM1-phosphorylatingactivity(D500fraction)with[13C]PEPandrecombinantPGAM1.
Similarreactionswithawhole-celllysateservedasapositivecontrol,anda[13C]PEPsamplethatcontainednocellularmaterialservedasaneg-ativecontrol.
WethenextractedmetabolitesfromtheresultingreactionstostudytheproductsderivedfromthelabeledPEPby[1H,13C]hetero-nuclearsingle-quantumcoherence(HSQC)nucle-armagneticresonance(NMR)(28).
Wedetected[13C]pyruvateinthewhole-celllysateasdeterminedbyanisolatedpeakcorrespondingtoa13C-labeledmethylgroupofpyruvate(29).
Nopyruvatewasobservedinthemock-treatedcontrol,indicatingthatPEPdidnotundergospontaneousdephos-phorylationandtautomerizationtopyruvateun-derthereactionconditions.
IncubationwiththeanionexchangefractioncontainingthePGAM1-phosphorylatingactivityalsocausedgenerationofpyruvate.
Theamountof[13C]pyruvatepro-ducedbytheD500fractionwasabout50%oftheamountproducedbyawhole-celllysate(Fig.
4D).
Thus,oneormorefactorsinthepartiallypurifiedfractionfromcelllysateslackingpyruvateki-nasemediatesPEP-dependentphosphorylationofPGAM1andconversionofPEPtopyruvate.
The[13C]pyruvatewasproducedfromPEPintheD500fractionatarateofabout30to60mM/min.
GiventhatthenumberofPGAM1moleculesinthisfractionissmallrelativetothenumberofPEPmoleculesconsumed,thisfractionmustalsocon-taintheabilitytoreleasePi.
TodeterminewhetherPiproductionfromPEPalsooccurredinthisfraction,weincubatedtheD500fractionwith[32P]PEPandrecombinantPGAM1andassessedthereleaseof32Piovertime(fig.
S13A).
WealsotestedwhethertherateofPiproductionwasen-hancedbyPGAM1.
AdditionofPGAM1shouldhavenoimpacton(orshoulddecrease)therateofPiproductionfromPEPifthisreactionisin-dependentofPEP-mediatedPGAM1phosphoryl-ation.
However,PGAM1additionstimulatedPiproductioninthefractionlackingpyruvatekinase(fig.
S13B),suggestingalinkbetweenPEP-dependentPGAM1phosphorylationandPEP-to-pyruvateconversionwithPirelease.
TheseresultssuggestthatreleaseofPifromeitherphosphoryl-atedPGAM1,PEP,orbothoccursinthisfractionandaccountsforhowPEPtopyruvateconversioncanoccurataratethatissuper-stoichiometrictotheamountofPGAM1present.
SelectivedetectionofPGAM1H11phosphoryl-ationandalteredglycolyticregulationinPKM2-expressingcellsandtumortissues.
TotestwhetherincreasedH11phosphorylationofPGAM1mightbecharacteristicofPKM2-expressingcellsasaconsequenceoftheirlowerpyruvatekinaseactivity,weengineeredH1299andA549lungcancercellstoexpresseitherPKM1orPKM2(4).
PGAM1expressionwassimilarregardlessofwhichpyruvatekinaseisoformwaspresent(fig.
S14A).
However,whentheisoelectricformsofPGAM1wereassessedby2DWesternblot,onlylysatespreparedfromPKM2-expressingcellshaddetectableamountsofthemostacidicspeciesthatcorrespondtoH11-phosphorylatedPGAM1(Fig.
5Aandfig.
S14B).
Thus,switchingcellsfromPKM2toPKM1expressionreducedtheamountofH11phosphorylatedPGAM1.
AccumulationofPEPincellsshouldenhancePGAM1phosphorylation.
BecausethePGAM1mutasereactioninvolvesa2,3-BPGinterme-diate(24),thisinturnshoulddriveconversionof3PGto2,3-BPG(fig.
S15).
Accordingly,acuteinhibitionofpyruvatekinaseincellswithPEP-dependentPGAM1phosphorylationactivityshouldincrease2,3-BPGlevels.
Totestthishy-pothesis,weacutelyinhibitedPKM2activityincellsbyadditionofpervanadatetoincreasepro-teintyrosinephosphorylation(5).
PervanadatehasnoeffectonPKM1activity(5);thus,comparingtheresponseofPKM2-expressingcellstoPKM1-expressingcellsseparatestheeffectsofacutePKM2inhibitionfromothereffectsofpervana-dateonmetabolism.
AcuteinhibitionofPKM2leadstoaboutatwofoldincreaseinPEPandayetlargerincreasein2,3-BPG(Fig.
5B),sug-gestingthatglycolysisinvolvingPGAM1phos-phorylationbyPEPoccursinPKM2-expressingcellsandthatrelativefluxthroughthisalterna-tivepathwayisincreasedwhenPKM2isinacti-vatedbyinteractionwithtyrosinephosphorylatedproteins.
Fig.
3.
TransferofthephosphateofPEPtoH11ofPGAM1.
(A)Recombinant6*His-taggedPGAM1(rPGAM1)wasphosphorylatedby[32P]PEPinacellex-tractandrecoveredbybindingtoNi-agarosebeads.
The[32P]rPGAMwasthendigestedwithtrypsin,andthepeptideswereseparatedbyusingHPLC.
Achro-matographidentifyingpeptidepeaksbyabsorbanceat208nmandthepres-enceof32Pdeterminedbyin-linescintillationcountingisshown.
Thepeptidepeakelutingat~26mincontaining32Pisdelineatedwithanarrow.
(B)TheHCDMS/MSspectrumforthephosphorylatedhistidine–containingpeptidepHGESAWNLENR(A,Ala;E,Glu;G,Gly;L,Leu;N,Asn;R,Arg;S,Ser;W,Trp)acquiredbyusingahybridLTQlineariontrap–OrbitrapXLmassspec-trometer(ThermoFisherScientific,SanJose,CA).
Thea1/pHisimmoniumionalongwiththeb-andy-seriesfragmentionsareallconsistentwiththesiteofphosphorylationlocalizedtotheHis1positionofthepeptide(H11inPGAM1).
PhosphatelossesobservedaretypicalofHisphosphorylation(21).
TheHis11phosphorylationsitewasconfirmedbyusingbothSequest(www.
thermofisher.
com/global/en/products/home.
asp)andMascot(www.
matrixscience.
com)databasesearchengineswithastatisticallysignificantexpectationvalueof0.
078.
(C)ExtractswerepreparedfromHEK293cellstransientlytransfectedwithN-terminallyFLAG-taggedPGAM1(Ctrl)orN-terminallyFLAG-taggedPGAM1whereH11wasmutatedtoN(H11N).
ExpressionofbothFLAG-taggedproteinsinrelationtoendogenousPGAM1wasdeterminedbyWesternblotusinganti-PGAM1.
Thesameextractswereincubatedwith[32P]PEP,andphosphorylationofPGAM1determinedbySDS-PAGEandautoradiography.
(D)rPGAM1wasincubatedwithacellextractcontaining[18O]phosphate-labeledPEPandnormalisotopic([16O]phosphate)ATPbeforerecoveryoftheH11-containingtrypticpeptidebyHPLCasdescribedin(A).
ThispeptidewasanalyzedbymicrocapillaryLC/MSusingthehighmassaccuracyoftheFT-MS–onlyscaninaLTQOrbitrap-XLmassspectrometerat30,000resolutionobtainingsub–2-parts-per-millionmassaccuracy.
Thepeaksatmass/charge(m/z)=697.
79,698.
79,and699.
79representthedoublychargedphosphorylatedpeptidepHGESAWNLENRthatisheavybytwo,four,andsixmassunitscorrespondingtotheincorporationofone,two,andthree18O-labeledoxygenatoms,re-spectively.
Thepeakatm/z=696.
79representsthephosphorylatedpeptidecontainingunlabeledoxygenatoms.
www.
sciencemag.
orgSCIENCEVOL32917SEPTEMBER20101497RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromFig.
5.
PhosphorylationofPGAM1H11incellsandtissuesexpressingPKM2.
(A)LysatesfromA549lungcancercellsengineeredtoexpresseitherPKM1orPKM2weresubjectedto2DIEFandSDS-PAGEandanalyzedbyWesternblotbyusinganti-PGAM1(aPGAM1)asshown.
Themostacidicspeciescor-respondingtoH11phosphorylationisindicatedwithanarrow.
(B)MetaboliteswereextractedfromH1299cellsengineeredtoexpresseitherPKM1orPKM2thatwereuntreatedortreatedwiththephosphataseinhibitorpervanadate(PV)for10mintoacutelyinhibitPKM2.
PKM2activityisdecreasedbyPVtreatment,whereasPKM1activityisnotchanged(5).
Thelevelsof2,3-BPGandPEPineachextractweredeterminedbymassspectrometry,andthechangesin2,3-BPGandPEPlevelsresultingfromPVtreatmentareshownforbothPKM1-andPKM2-expressingcells.
(C)Prostatetissuewasremovedfrom12-week-oldmiceharboringaconditionalalleleofthePtentumorsuppressorgenethatalsodid(Ptenpc/)ordidnot(Ptenpc+/+)containatransgenetoexpressCrerecombinaseinthepros-tatetodeletePten.
ThePtenpc/wasconfirmedtohavehigh-gradeprostateneoplasiabyhistology.
TheexpressionofPKM1orPKM2ineachtissuewasdeterminedbyWesternblotasshown.
(D)Prostatetissuelysatesfromthesamemicedescribedin(C)weresubjectedto2DIEFandSDS-PAGEandanalyzedbyWesternblotusinganti-PGAM1asshown.
Themostacidicspeciescorre-spondingtoH11phosphorylationisindicatedwithanarrow.
(E)Abreasttumor(cancer)wasremovedfroma9-month-oldmouseharboringaconditionalalleleoftheBrca1tumorsuppressorgeneandatransgenetoexpressCrerecombinaseinthebreasttodeleteBrca1.
NormalbreasttissuewasremovedfromamousenotexpressingCreandhencewhereBrca1wasnotdeletedinthebreast.
NormalbreastexpressesPKM1;breasttumorsexpressPKM2(4)(fig.
S14C).
Lysatesfromthenormalbreasttissueandthebreasttumorweresubjectedto2DIEFandSDS-PAGEandanalyzedbyWesternblotusinganti-PGAM1asshown.
ThemostacidicspeciescorrespondingtoH11phosphorylationisindicatedwithanarrow.
AIEF+-PKM1PKM2PGAM1DPtenpc+/+Ptenpc-/-PKM1PKM2+-IEFPGAM1Ptenpc+/+Ptenpc-/-+-IEFPGAM1NormalbreastBreastcancerCE051015202530354045PKM1PKM2FoldChangewithPVFoldChangewithPV2,3-BPG012345678910PKM1PKM2PEPBFig.
4.
AssociationofPGAM1phosphorylationwithconversionofPEPintopyruvateintheabsenceofpyruvatekinase.
(A)rPGAM1wasaddedtoaHEK293cellextractintheabsence(Ctrl)orpresenceofPEP(+PEP).
Thereactionswereanalyzedbyusing2DIEFandSDS-PAGEfollowedbyWesternblotusinganti-PGAM1.
Thenewlyresolved,moreacidicspeciespresentonlyinthePEP-containingreactionarein-dicatedbyanarrow.
(B)AHEK293celllysatewascentrifugedat100,000g,andtheS100supernatantfractionatedoveraweakanionexchange(DEAE)column.
TheFTandfractionselutedsequentiallywith100mM,200mM,and500mMNaClwerecollectedandincubatedwithrPGAM1and[32P]PEP.
TheabilityofeachfractiontophosphorylatePGAM1wasdeterminedbySDS-PAGEandautoradiography.
Theamountofenolaseandpyruvatekinase(PK)ineachfractionwasdeterminedbyWesternblot.
(C)TheenolaseactivitywasdeterminedintheFTand500mMNaCl(D500)fractionsdescribedin(B).
Inaddition,theADP-dependentpyruvatekinaseactiv-ityineachfractionwasdetermined.
(D)The2,3-[13C]PEPwasincubatedwithaHEK293cellS100fraction(Celllysate)orthe500mMNaClfractiondescribedin(B)(D500),whichcontainedthePGAM1-phosphorylatingactivity.
The[13C]PEPwasalsoincubatedundertheidenticalreactionconditionsintheabsenceofanyprotein(Ctrl).
Quantificationoftheconversionof[13C]PEPto[13C]pyruvatewasmea-suredbyintegratingtheintensityofthepyruvatepeakanddividingbytheintensityoftheinternalstandardconsistingof2mM4,4-dimethyl-4-silapentane-1-sulfonicacidforeach[1H,13C]HSQCspectracollected.
Thisratioisgraphedforeachcondition.
ArPGAM1FT100200500DEAE32P-PEPEnolasePKBCDCtrl+PEPrPGAM1PGAMrPGAM1PGAMIEF+-PGAM1-50050100150200250FTD500Pkactivity(pmol/s/gequivprot)PKactivity-50510152025303540FTD500Enolaseactivity(pmol/s/gequivprot)Enolaseactivity012345CelllysateD500CtrlIpyruvateIstandard17SEPTEMBER2010VOL329SCIENCEwww.
sciencemag.
org1498RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfromTumorsexpressPKM2,whereasPKM1isexpressedinmanynormaltissues(3,4).
Tode-terminewhetheracorrelationbetweenPGAM1H11phosphorylationstatusandPKM2expres-sioncouldalsobeobservedforcancersinvivo,weanalyzedtissuesfromanimalsby2DWesternblotwithanantibodytoPGAM.
PKM1wasexpressedinnormalprostatetissue,whereasneoplasticprostatetissueexpressedPKM2(Fig.
5C).
Analysisofthesesamecelllysatesby2DWesternblotwithanantibodyagainstPGAM1(anti-PGAM1)revealedthattheisoelectricspeciescorrespondingtoH11-phosphorylatedPGAM1wasonlydetectableintheneoplasticprostatetissue(Fig.
5D).
Inamousewithbreast-specificdeletionoftheBrca1tumorsuppressorgene,breasttumortissuehadanevenmoredramaticshiftinisoelectricmigrationofPGAM1associatedwiththeH11-phosphorylatedspecies(Fig.
5E),andthiscorrelatedwithexpressionofPKM2(fig.
S14C).
Thus,PEP-dependentphosphoryl-ationofH11onPGAM1likelyoccursintumorsaswellasincelllinesexpressingPKM2.
Discussion.
Wehaveshownthatthehigh-energyphosphateofPEPcanbetransferredtothecatalytichistidine(H11)ofPGAM1byanenzymaticprocessthatdoesnotrequireenolase-dependentconversionto2PG.
PhosphorylationofH11isknowntoberequiredforPGAM1catalyticfunction,and2,3-BPGhasbeencharac-terizedasthecofactorrequiredforPGAM1acti-vation(17).
OurresultsrevealthatPGAM1canalsobeactivatedbyPEP.
Theactivitythatcat-alyzesphosphatetransferfromPEPtoPGAM1isseparablefromthewell-knownPEP-metabolizingenzymes,pyruvatekinaseandenolase,andtheby-productofthereactionappearstobepyruvate.
Importantly,wefindthat,incellsexpressingPKM2,asignificantfractionofPGAM1mi-gratesatanisoelectricpointconsistentwiththephospho-H11speciesandthatreplacingPKM2withthemoreactivePKM1isoformresultsindisappearanceofthisspecies.
TheseresultsareconsistentwithamodelwherePKM2-expressingcellsuseagreaterfractionofPEPforchargingPGAM1andlessforthesynthesisofATP.
DespitethefactthatPGAM1hasnotbeenconsideredarate-limitingenzymeinglycolysis,thedifferentialH11phosphorylationofPGAM1weobservedinPKM2-versusPKM1-expressingcellsandtissuessuggeststhatthisenzymemayhaveapreviouslyunappreciatedregulatoryfunc-tionincontrollingglycolysisinproliferatingcells.
PGAM1isuniqueamongtheglycolyticenzymesinthatitstranscriptionisregulatedbythetumorsuppressorp53(30),andincreasedexpressionofPGAM1hasbeenreportedtoimmortalizeprimarycellsthroughanunknownmechanism(31).
PGAM1wasalsoidentifiedasthetargetofacompoundfromachemicalgenomicsscreenformoleculesthatinhibitbreastcancercellgrowth(32).
Thus,oneimportantconsequenceofdown-regulatingPKM2activitybytyrosinekinasesmaybetoincreaseH11-phosphorylatedPGAM1.
Phospho-rylationofH11onPGAM1increasesthemutasefunctionoftheenzyme.
ThisgeneratesapositivefeedbackloopsuchthatproductionofPEPin-creasestheenzymaticactivityofPGAM1.
OnepossibilityisthatthisfeedbackloopmaypromotetheredistributionofmetabolitesupstreamofPGAM1intobiosyntheticpathwaysthatbranchfromglycolysis.
Weproposethat,inadditiontopyruvatekinase,anotheractivitytoconvertPEPintopyruvatemaybeactiveincells(fig.
S15).
Theexistenceofsuchanalternateglycolyticpathwaymayexplainhowcancercellswithlesspyruvatekinaseactivitycontinuetodisplayahighrateofglycolysis.
TherateofPEPtopyruvateconversionobservedintheabsenceofpyruvatekinaseiscomparabletothemaximumenzymevelocity(Vmax)ofpyru-vatekinase[65mM/min(33)],suggestingthatthisactivitycouldaccountforasignificantamountofthepyruvateproducedfromglucoseincellswithalessactiveformofpyruvatekinase.
Whencatalyzedbypyruvatekinase,theconver-sionofPEPintopyruvateiscoupledtoATPgeneration(34).
PhosphorylationofPGAM1byPEPdoesnotdirectlygenerateATPbutgeneratespyruvate.
Inorderforasignificantamountofpyruvatetobegeneratedbythisalternativepathway,thephosphohistidineofPGAM1mustturnover.
Althoughconversionof3PGto2PGdoesnotresultinnetlossofphosphohistidine,spontaneoushydrolysisofphosphohistidineonPGAM1doesoccur(17).
Also,2,3-BPGcanbeproducedbyadditionof3PGor2PGtophosphorylated-PGAM1(reversalofthe2,3-BPGchargingreaction),andtheresulting2,3-BPGcanbehydrolyzedto2PGandPi(35).
Lastly,itispossiblethattheactivityresponsibleforPGAM1phosphorylationcanalsoactasaPEPphosphatase.
Eachofthesepossibilitiesre-sultsinthenetconversionofPEPtopyruvateandPiwithnoATPsynthesis.
ThislackofATPsyn-thesismayallowcellstometabolizeglucosebyamodifiedglycolysisthatdoesnotgenerateATPandprovidesanadvantagetoproliferatingcells.
Historically,effortstounderstandaerobicgly-colysisstressedtheimportanceofATPconsump-tiontoallowthehighrateofglucosemetabolismobservedintumorcells(36).
CellsmustavoidATPproductioninexcessofdemandtoavoidallostericinhibitionofphosphofructokinaseandotherrate-limitingstepsinglycolysisthatarein-hibitedbyahighATP/adenosinemonophosphateratio(34).
Thus,inhibitionofPKM2bycellgrowthsignalsmayservetouncoupletheabilityofcellstoassimilatenutrientsintobiosyntheticpathwaysfromtheproductionofATPandac-countforwhyPKM2activityisdecreasedinrapidlydividingcells.
ReferencesandNotes1.
R.
G.
Jones,C.
B.
Thompson,GenesDev.
23,537(2009).
2.
M.
G.
VanderHeiden,L.
C.
Cantley,C.
B.
Thompson,Science324,1029(2009).
3.
S.
Mazurek,C.
B.
Boschek,F.
Hugo,E.
Eigenbrodt,Semin.
CancerBiol.
15,300(2005).
4.
H.
R.
Christofketal.
,Nature452,230(2008).
5.
H.
R.
Christofk,M.
G.
VanderHeiden,N.
Wu,J.
M.
Asara,L.
C.
Cantley,Nature452,181(2008).
6.
T.
Hitosugietal.
,Sci.
Signal.
2,ra73(2009).
7.
MaterialsandmethodsareavailableassupportingmaterialonScienceOnline.
8.
J.
D.
Dombrauckas,B.
D.
Santarsiero,A.
D.
Mesecar,Biochemistry44,9417(2005).
9.
M.
H.
SaierJr.
,J.
Reizer,Mol.
Microbiol.
13,755(1994).
10.
J.
Deutscher,C.
Francke,P.
W.
Postma,Microbiol.
Mol.
Biol.
Rev.
70,939(2006).
11.
J.
N.
Burnell,M.
D.
Hatch,Arch.
Biochem.
Biophys.
231,175(1984).
12.
J.
N.
Burnell,M.
D.
Hatch,Arch.
Biochem.
Biophys.
245,297(1986).
13.
S.
Minakami,H.
Yoshikawa,Biochem.
Biophys.
Res.
Commun.
18,345(1965).
14.
R.
Lagunas,C.
Gancedo,Eur.
J.
Biochem.
137,479(1983).
15.
S.
Klumpp,J.
Krieglstein,Eur.
J.
Biochem.
269,1067(2002).
16.
P.
G.
Besant,P.
V.
Attwood,Biochim.
Biophys.
Acta1754,281(2005).
17.
L.
A.
Fothergill-Gilmore,H.
C.
Watson,Adv.
Enzymol.
Relat.
AreasMol.
Biol.
62,227(1989).
18.
M.
E.
Lee,T.
Nowak,Biochemistry31,2172(1992).
19.
J.
V.
Olsenetal.
,Nat.
Methods4,709(2007).
20.
K.
F.
Medzihradszky,N.
J.
Phillipps,L.
Senderowicz,P.
Wang,C.
W.
Turck,ProteinSci.
6,1405(1997).
21.
A.
J.
Kleinnijenhuis,F.
Kjeldsen,B.
Kallipolitis,K.
F.
Haselmann,O.
N.
Jensen,Anal.
Chem.
79,7450(2007).
22.
U.
M.
Hohenester,K.
Ludwig,J.
Krieglstein,S.
Konig,Anal.
Bioanal.
Chem.
397,3209(2010).
23.
X.
L.
Zu,P.
G.
Besant,A.
Imhof,P.
V.
Attwood,AminoAcids32,347(2007).
24.
D.
N.
Perkins,D.
J.
Pappin,D.
M.
Creasy,J.
S.
Cotrell,Electrophoresis20,3551(1999).
25.
J.
K.
Eng,A.
L.
McCormack,J.
R.
YatesIII,J.
Am.
Soc.
MassSpectrom.
5,976(1994).
26.
W.
A.
Blttler,J.
R.
Knowles,Biochemistry19,738(1980).
27.
M.
Engel,S.
Mazurek,E.
Eigenbrodt,C.
Welter,J.
Biol.
Chem.
279,35803(2004).
28.
L.
E.
Kay,P.
Keifer,T.
Saarinen,J.
Am.
Chem.
Soc.
114,10663(1992).
29.
D.
S.
Wishartetal.
,NucleicAcidsRes.
35,D521(2007).
30.
E.
C.
Cheung,K.
H.
Vousden,Curr.
Opin.
CellBiol.
,inpress;publishedonline8January2010.
31.
H.
Kondohetal.
,CancerRes.
65,177(2005).
32.
M.
J.
Evans,A.
Saghatelian,E.
J.
Sorensen,B.
F.
Cravatt,Nat.
Biotechnol.
23,1303(2005).
33.
S.
Ainsworth,N.
MacFarlane,Biochem.
J.
131,223(1973).
34.
A.
L.
Lehninger,D.
L.
Nelson,M.
M.
Cox,PrinciplesofBiochemistry(Worth,NewYork,ed.
2,1993).
35.
J.
Cho,J.
S.
King,X.
Qian,A.
J.
Harwood,S.
B.
Shears,Proc.
Natl.
Acad.
Sci.
U.
S.
A.
105,5998(2008).
36.
E.
Racker,J.
Cell.
Physiol.
89,697(1976).
37.
WethankA.
CarrecedoandP.
P.
Pandolfifornormalandneoplasticprostatetissue;L.
BurgaandG.
Wulffornormalbreastandbreasttumor;G.
Bellingerforassistancewithmousedissection;M.
Sasakiforhelpgeneratingconstructs;V.
Vyas,A.
Subtelny,L.
Freimark,X.
Yang,L.
Schmidt,andM.
Balastikfortechnicalassistance;M.
BayroandS.
HybertsforhelpwithprocessingtheNMRspectra;andA.
Shaywitz,J.
Hutti,C.
Benes,D.
Anastasiou,A.
Couvillion,andA.
Saciforhelpfuldiscussions.
ThisworkwaspartiallysupportedbytheDamonRunyonCancerResearchFoundation(M.
V.
H.
),theBurroughsWellcomeFund(M.
V.
H.
),theAmericanCancerSociety(J.
W.
L.
),Dana-Farber/HarvardCancerCenter(J.
M.
A.
andJ.
W.
L.
),andbygrantsfromtheNIH(1K08CA136983toM.
V.
H.
,5P30CA006516-43toJ.
M.
A.
,5T32CA009361-28toJ.
W.
L.
,R21/R33DK070299andP01GM047467toG.
W.
,R01AI078063andR21CA128620toJ.
D.
R.
,R01-GM56302andP01CA089021toL.
C.
C.
,and1P01CA120964-01AtoJ.
M.
A.
andL.
C.
C.
).
SupportingOnlineMaterialwww.
sciencemag.
org/cgi/content/full/329/5998/1492/DC1MaterialsandMethodsFigs.
S1toS15TableS1References5February2010;accepted27July201010.
1126/science.
1188015www.
sciencemag.
orgSCIENCEVOL32917SEPTEMBER20101499RESEARCHARTICLESonSeptember20,2010www.
sciencemag.
orgDownloadedfrom