磁共振,极化

MeasuringintracellularpHintheheartusing

hyperpolarizedcarbondioxideandbicarbonate:a1331CandPmagneticresonancespectroscopystudy

MarieA.Schroeder1*,PawelSwietach2,HelenJ.Atherton1,FerdiaA.Gallagher3,4,PhillipLee1,GeorgeK.Radda1,KieranClarke1,andDamianJ.Tyler1

CardiacMetabolismResearchGroup,DepartmentofPhysiology,AnatomyandGenetics,UniversityofOxford,SherringtonBuilding,ParksRoad,OxfordOX13PT,UK;2ProtonTransportGroup,DepartmentofPhysiology,AnatomyandGenetics,UniversityofOxford,SherringtonBuilding,ParksRoad,OxfordOX13PT,UK;3DepartmentofRadiology,UniversityofCambridge,Addenbrooke’sHospital,CambridgeCB22QQ,UK;and4CRUKCambridgeResearchInstitute,CambridgeCB20RE,UKReceived16October2009;revised29November2009;accepted10December2009;onlinepublish-ahead-of-print15December2009Timeforprimaryreview:16days

1Aims

TechnologicallimitationshaverestrictedinvivoassessmentofintracellularpH(pHi)inthemyocardium.Theaimofthisstudywastoevaluatethepotentialofhyperpolarized[1-13C]pyruvate,coupledwith13Cmagneticresonancespectroscopy(MRS),tomeasurepHiinthehealthyanddiseasedheart.

.....................................................................................................................................................................................MethodsHyperpolarized[1-13C]pyruvatewasinfusedintoisolatedratheartsbeforeandimmediatelyafterischaemia,andthe

13formationof13CO2andH13CO2CMRS.TheHCO2andresults3wasmonitoredusing3/CO2ratiowasusedintheHender-son–HasselbalchequationtoestimatepHi.Wetestedthevalidityofthisapproachbycomparing13C-basedpHimeasurementswith31PMRSmeasurementsofpHi.TherewasgoodagreementbetweenthepHimeasuredusing13Cand31PMRSincontrolhearts,being7.12+0.10and7.07+0.02,respectively.Inreperfusedhearts,13Cand31PmeasurementsofpHialsoagreed,although13Cequilibrationlimitedobservationofmyocardialrecoveryfromacidosis.Inheartspre-treatedwiththecarbonicanhydrase(CA)inhibitor,6-ethoxyzolamide,the13Cmeasurementunderestimatedthe31P-measuredpHiby0.80pHunits.Mathematicalmodellingpredictedthatthevalidityofmeasur-13ingpHifromtheH13CO23/CO2ratiodependedonCAactivity,andmaygiveanincorrectmeasureofpHiunderconditionsinwhichCAwasinhibited,suchasinacidosis.Hyperpolarized[1-13C]pyruvatewasalsoinfusedinto

13healthylivingrats,whereinvivopHifromtheH13CO23/CO2ratiowasmeasuredtobe7.20+0.03.

.....................................................................................................................................................................................ConclusionMetabolicallygenerated13CO2andH13CO23canbeusedasamarkerofcardiacpHiinvivo,providedthatCAactivity

isatnormallevels.

-----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords

Magneticresonancespectroscopy†Hyperpolarization†pH†Ischaemia†Carbonicanhydrase

1.Introduction

Therapidonsetofacidosisisawell-documentedcharacteristicofmyo-cardialischaemia.1,2Underpoorcoronaryperfusion,anaerobicglycolysisincreasesintheheart,producingintracellularprotonsandlacticacidthataccumulateintheintra-andextracellularspaces.3SomeoftheacidreactswithHCO23toformCO2,whichaddstoanyCO2generatedbyresidualoxidativemetabolism.Accumulationofprotons,lacticacid,andCO2intheischaemicheartdecreasesintracellularpH(pHi)fromnormallevelsofaround7.1–7.2.1,2Transientacidosisduringischaemiamaybebeneficial,asitdecreasesthemajoradenosinetriphosphate(ATP)con-sumer,contractility,andthusconservesATPforiontransport.4However,theATPreductioncausedbysevereandsustainedischaemiadecreasesNaþ,Kþ-ATPaseactivity,whichincreasesmyocardialNaþlevels.IncreasedNaþinhibitsCa2þextrusionviatheNaþ/Ca2þexchan-ger,thuselevatingmyocardialCa2þanddamagingthemyocardium.5*Correspondingauthor.Tel:þ441865282249;fax:þ441865282272,Email:marie.schroeder@dpag.ox.ac.uk

PublishedonbehalfoftheEuropeanSocietyofCardiology.Allrightsreserved.&TheAuthor2009.Forpermissionspleaseemail:journals.permissions@oxfordjournals.org.

Theonlineversionofthisarticlehasbeenpublishedunderanopenaccessmodel.Usersareentitledtouse,reproduce,disseminate,ordisplaytheopenaccessversionofthisarticlefornon-commercialpurposesprovidedthattheoriginalauthorshipisproperlyandfullyattributed;theJournal,LearnedSocietyandOxfordUniversityPressareattributedastheoriginalplaceofpublicationwithcorrectcitationdetailsgiven;ifanarticleissubsequentlyreproducedordisseminatednotinitsentiretybutonlyinpartorasaderivativeworkthismustbeclearlyindicated.Forcommercialre-use,pleasecontactjournals.permissions@oxfordjournals.org.

MeasuringcardiacintracellularpHusinghyperpolarizedmetabolites

83

Figure1Themetabolicfateofinfusedhyperpolarized[1-13C]pyruvateisshown.Infusedpyruvateisoxidizedtoformacetyl-CoAandthe

by-product,13CO2,inthemitochondria,bytheenzymePDH.Mitochondrial13CO2isthenthoughttorapidlydiffuseintothecytosol,andsub-sequentlyoutofthecell.Intheory,13CO2ineitherthemitochondriaorthecytosolcouldequilibratewithH13CO23,mediatedbyCAactivity.

13132OnceCO2diffusesintothebloodstream,itwillrapidlyequilibratewithHCO3viaCA.CA,carbonicanhydrase;PDH,pyruvatedehydrogenase;MCT,mono-carboxylatetransporter;NHE,sodiumprotonexchanger;NBC,sodiumbicarbonatecarrier;CBE,chloridebicarbonateexchanger;NCE,sodiumcalciumexchanger.

Pmagneticresonancespectroscopy(MRS)haslongbeenthegoldstandardforpHimeasurementintheisolatedperfusedheart,6basedonthechemicalshiftoftheinorganicphosphate(Pi)peak.7,8However,31PMRScannotmeasurecardiacpHiinvivo,because2,3-diphosphoglycerate(2,3-DPG)intheventricularbloodcontami-natesthemyocardialPipeak.6,9,10ThepH-dependentequilibriumbetweenbicarbonateandCO2hasbeenusedtomeasureextracellularpH(pHo)non-invasivelyintumours.11Byinfusinghyperpolarized13C-bicarbonateintravenously,MRwasusedtoimagethedistributionofhyperpolarizedbicarbonateandCO2andapHmapwasgeneratedusingtheHenderson–Hassel-balchequation:

󰀂󰀃À½HCO3󰀃

pH¼pKaþlog

½CO2󰀃

ð1Þ

31wherepKaistheacid-dissociationconstantofCO2,whichis6.15intheKrebs–Henseleitbuffer.12ForcorrectapplicationoftheHenderson–Hasselbalchequation,thefollowingtwoconditionsmustbemet:

CO2toH13CO23exchangekinetics,catalysedbycarbonicanhy-drase(CA),mustberapidand13(ii)CO2andH13CO23signalsmustbedetectedsimultaneously

fromthesamecellularcompartment.(i)

13Intumours,theHenderson–Hasselbalchequationwasappliedcor-rectlybecauseofthehighCAactivityonthesurfaceoftumourcells13andwithinerythrocytes,14andtheslow,transporter-mediated,cellularuptakeofinfusedbicarbonate.15AsimilarapproachmaybeusefulformeasuringpHiintheinvivoheart.16,17Infusionofhyperpolarized[1-13C]pyruvateresultsinmito-chondrialproductionofhyperpolarized13CO2bypyruvatedehydro-genase(PDH,Figure1).18Hyperpolarizationbythedynamicnuclearpolarizationmethodincreasesthe13CMRsensitivityofpyruvate,andother13C-labelledmetabolites,morethan20000-fold.19Further,whenahyperpolarizedmetaboliteisinfusedintotissue,thehigh-sensitivity13Clabelistransferredtothetracer’smetabolicproducts,enablingunprecedentedreal-timevisualizationofthebio-chemicalmechanismsofnormalandabnormalmetabolism.Onlymetabolicprocessesthatoccurrapidlycanbemonitoredusinghyper-polarized13CMRmethodsbecausethehyperpolarizedsignaldecaystothermalequilibriumaccordingtoitsinherentspin–latticerelax-ationtime(inthecaseof[1-13C]pyruvatewithatimeconstantof50–60s).Intheory,simultaneousdetectionofhyperpolarized1313132[1-C]pyruvate-derivedCO2andHCO3couldbeusedtomeasurepHi.However,incardiacmyocytes,theconditionsrequiredforthecorrectuseoftheHenderson–Hasselbalchequationmaynotapply.AsshowninFigure1,metabolicallygeneratedCO2diffusesrapidlyfromitssiteofproductionintothecytosol,andsubsequentlyintotheextracellularspace.20CardiacmyocytesalsohaveHCO23þtransportactivity,throughproteinssuchastheNa–HCO232212co-transporterandtheCl/HCO3exchanger.StudiesofCAlocal-izationandkineticsincardiacmyocytessuggestlowintracellularCAactivity.21Undertheacidicconditions,typicalofischaemia,CAactivityisexpectedtobeevenlower.22,23Further,PDHfluxpost-ischaemiamustbesufficientlyhightoenableMRdetectionof13CO2and

13H13CO2CMRsignal.3priortodecayofthehyperpolarized,84Theaimofthepresentstudywastoevaluatethepotentialofhyperpolarized13CMRforthenon-invasivemeasurementofpHiintheheart.WemeasuredPDHfluxandtheproductionof13CO2andH13CO23inisolatedheartsbeforeandafterischaemiaandwithCAactivityinhibited.Weusedmathematicalmodellingtodetermine

whether,andunderwhatconditions,theH13CO23/13CO2ratiocouldbeusedtomeasurecardiacpHi.Finally,wemeasuredpHiintheinvivoratheart.

2.Methods

2.1Theisolatedperfusedratheart

AllinvestigationsconformedtotheGuidefortheCareandUseofLab-oratoryAnimalspublishedbytheUSNationalInstitutesofHealth(NIHPublicationNo.85-23,revised1996),theHomeOfficeGuidanceontheOperationoftheAnimals(ScientificProcedures)Act,1986(HMSO),andtoinstitutionalguidelines.MaleWistarrats(󰀂300g)wereanaesthetizedusinga0.7mLipinjectionofpentobarbitalsodium(200mg/mLEuthatal).Thebeatingheartswerequicklyremovedandarrestedintheice-coldKrebs–Henseleitperfusionbuffer,andtheaortawascannulatedforperfusioninrecirculatingretrogradeLangendorffmodeataconstant85mmHgpressureand378Ctemperature.24TheKrebs–Henseleitbicarbonateperfusionbuffercontained1.2mMinor-ganicphosphate(KH2PO4),11mMglucose,and2.5mMpyruvateandwasaeratedwithamixtureof95%oxygen(O2)and5%carbondioxide(CO2)togiveafinalpHof7.4at378C.Thebroad-spectrumCAinhibitor,6-ethoxyzolamide(ETZ),wasdissolvedindimethylsulfoxide(DMSO)andaddedtotheperfusatetoachieveafinalconcentrationof100mM(withDMSO,0.01%oftotalbuffervolume).ETZwasexpectedtoevenlydis-tributethroughouttheintra-andextracellularspacestoinhibitallcardiacCAisoforms.20Unlessspecified,compoundswereobtainedfromSigma(Gillingham,UK).ForfurtherdetailsoftheLangendorffheartperfusionmethod,seeSupplementarymaterialonline,S1.

2.2Experimentalprotocols

2.2.1Controlprotocol

Isolatedhearts(n¼6)wereperfusedfor󰀂30minat85mmHg.Fortheinitial20min,31PMRSwasusedtomeasurepHi.Afterthis,hyperpol-arized[1-13C]pyruvatewasinfusedandtheprogressof13C-labelledcom-poundswasfollowedusing13CMRS.

2.2.2InhibitionofCA

Hearts(n¼5)wereperfusedfor󰀂30min.After10minofperfusioninnormalbuffer,theheartswereswitchedtobuffercontaining100mMETZ.31PMRSwasperformedfor20min(10minbeforeand10minafterswitchovertoETZ-containingbuffer).TenminutesafterthestartofETZperfusion,hyperpolarized[1-13C]pyruvatewasinfusedandMRSwasswitchedfrom31Pto13C.

2.2.3Reperfusionfollowingischaemia

Hearts(n¼12)wereperfusedfor󰀂30min,followedby10minoftotal,globalischaemia,and15minreperfusion.31PMRSspectrawereacquiredfor20minimmediatelybeforeand9minduringischaemia.Hyperpol-arized[1-13C]pyruvatewasinfusedimmediatelyafterischaemia,suchthatheartswerereperfusedwithhyperpolarizedtracer.Insomehearts(n¼6),31PMRSwasperformedthroughoutthereperfusionperiod.Inotherhearts(n¼6),13CMRSwasperformedfor2minduringinitialreperfusionwithhyperpolarized[1-13C]pyruvate,followedby10minof31PMRspectralacquisition.Fordetailsofthe[1-13C]pyruvatepreparationanddelivery,seeSupplementarymaterialonline,S2.

M.A.Schroederetal.2.3Magneticresonancespectroscopy31PMRspectrawereacquiredat202.5MHzusinga308radiofrequency(RF)pulseandarepetitiondelayof0.25s.Thephosphocreatine(PCr)resonancewassetat0ppmandthechemicalshiftsofallpeakswerereferencedtothatofPCr.Eachspectrumconsistedof120transients,givingatotalacquisitiontimeof30s.Asthesepartiallysaturatedspectrahadshorterrepetitiontimesthanthelongitudinalrelaxationtimeof31Pnuclei,anunsaturatedspectrumwasinitiallyacquiredfromtheheartsusinga908pulsewithrepetitiontimeof15sand40transients,andanacquisitiontimeof10min.Theunsaturatedspectrawereusedtocorrectmetaboliteconcentrationsfortheeffectsofsaturation.

Acquisitionof13CMRspectracommencedimmediatelyafterinfusionofhyperpolarized[1-13C]pyruvateand[1-13C]pyruvateinfusioncontinuedthroughoutacquisition.Spectrawereacquiredwith1stemporalresolutionover2min(excitationflipangle¼308,120acquisitions).Spectrawerecentredat150ppmandreferencedtothe[1-13C]pyruvateresonanceat171ppm,and4096pointswereacquiredoverabandwidthof100ppm.Invivohyperpolarized[1-13C]pyruvateMRSexperimentswereper-formedasdescribedpreviously.18Briefly,[1-13C]pyruvicacidwashyper-polarizedanddissolved/neutralizedinaprototype13Cpolarisersystem.25Eachlivingrat(n¼6)waspositionedattheisocentreofa7TVarianhori-zontalboreMRscanner,withadual-tuned1H/13Ccoillocalizedovertheanimal’schest.Aqueoushyperpolarized[1-13C]pyruvate(80mmol)wastheninfusedintoalivingratviathetailveinover10s,andcardiac13Cspectrawereacquiredwithalow7.58flipangleeverysecondfor1min.Forfurtherdetailsofinvivohyperpolarized[1-13C]pyruvateMRSexper-iments,refertoSupplementarymaterialonline,S3.

2.4Dataanalysis

2.4.1Carbon-13

Cardiac13CMRspectrawereanalysedusingtheAMARESalgorithm,asimplementedinthejMRUIsoftwarepackage.26SpectrawereDCoffsetcor-rectedbasedonthelasthalfofacquiredpointsandpeakscorrespondingwith[1-13C]pyruvateanditsmetabolicderivativeswerefittedassumingaLorent-zianlineshape,initialpeakfrequencies,relativephases,andlinewidths.Forspectraacquiredfromperfusedrathearts,themaximumpeakareaofeachmetaboliteoverthe2minofacquisitionwasdeterminedforeachseriesofspectraandexpressedasapercentageofthemaximum[1-13C]pyruvateresonance.18Therateofsignalproductionforeachmetab-olite,inpercentpersecond(%/s),wasmeasuredastheslopeofthemeanmetaboliteincreaseoverthefirst5sfollowingitsappearance,overwhichtimethemetabolitesignalincreasedlinearly.Additionally,afirst-orderexponentialsignaldecaytermwasfittoeachmetabolitepeakfromthepointofmaximumsignaloverthecourseofsignaldecay.Decayofthehyperpolarizedsignaldependsontheintrinsicspin–latticerelaxationofthenucleus,productionandconsumptionratesofthemetabolite,andmetabolitewashout,andmaythereforeprovideinformationaboutmetab-oliteaccumulationinthestatesofno-flowischaemiaandCAinhibition.

AveragetimecoursesforH13CO23,

13CO2,andtheirsumwerecalcu-latedforallheartsforfurtherdataanalysis.H13CO23plus

13CO2,normal-izedtothemaximumpyruvatepeakareatoallowforanydifferencesinpolarization,wasusedasaqualitativeindicatorofPDHflux.The

averageH13CO23and

13CO2timecourseswereinsertedintoanappliedformoftheHenderson–Hasselbalchequation:

15þlog

󰀂

½H13CO3ÀpR¼6:󰀃

󰀃

½13COð2Þ

2󰀃

TheoutputofEq.(2)isavariablepRwhichshould,underthetwocon-ditionsoutlinedinSection1,measurepH.Upontheinitialarrivalof[1-13C]pyruvate,therelativeproportionsof13CO2andH13CO23(andthuspR)equilibratedoverseveralsecondstoreacha

MeasuringcardiacintracellularpHusinghyperpolarizedmetabolitessteady-statevalue.ThecalculatedpRwasfittoafirst-orderexponentialequationtodeterminethesteady-statevalueandtimeconstant.2.4.2Phosphorus-31Cardiac31PMRspectrawereanalysedusingtheAMARESalgorithminthejMRUIsoftwarepackage.26SpectrawerecorrectedforDCoffsetusingthelasthalfofacquiredpoints.ThePCr,Pi,a-,b-,andg-ATPresonanceswerefittedassumingaLorentzianlineshape,peakfrequencies,relativephases,linewidths,andJ-couplingparameters.pHiwascalculatedfromtheP31ichemi-calshift.8,27AbsolutePmetaboliteconcentrationswerecalculatedusinganATPconcentrationof10.6mMfromthefirstg-ATPpeakarea28andexpressingallotherATPandPCrpeakareasrelativetothisarea.272.4.3ModellingAsystemofordinarydifferentialequationswasformulatedtotestthesuit-abilityofusingtheCO2–HCO23equilibriumtomeasurepHi.Fordetailsofthemathematicalmodel,seeSupplementarymaterialonline,S4.2.4.4StatisticalmethodsDataaregivenasmean+standarderror.Statisticalsignificancesbetweenpre-andpost-ischaemicgroups,andpre-ischaemicandETZ-perfusedgroups,wereassessedusingapairedStudent’st-test.Statisticalsignifi-cancewasconsideredatP,0.05.3.Results3.1MyocardialenergeticsCardiacfunctionand31PMRspectrawerecharacteristicoftheiso-latedratheartduringpre-ischaemia,ischaemia,andreperfusion.1Adescriptionofcardiacfunctionthroughouttheprotocolandanexampleofa31PspectrumofaheartbeforeischaemiaareshowninSupplementarymaterialonline,S5.Pre-ischaemia,theaverage[ATP]was10.6+0.7mMand[PCr]was19.7+0.9mM(Figure2).Twominutesafterstoppingcoronaryflow,[PCr]decreasedto3.2mM,toremainat1.1–2.1mMfortheremainderofischaemia.Therateof[ATP]hydrolysisduringischaemiawas0.14+0.10mM/min.Fiveminutesafterreperfusion,PCrhadrecoveredto17.6+1.9mM,whereasATPremainedat8.2+2.5mM.PerfusionwithETZhadnoeffecton[ATP]or[PCr]throughouttheperfusionprotocol(datanotshown).PriortoETZperfusion,heartshadanaveragePCrof17.8+1.9mMandATPof10.6+0.5mM.DuringETZperfusion,theaverage[PCr]was18.0+1.5mMand[ATP]was10.3+0.9mM.

3.2PDHflux

Arepresentativespectrumof[1-13C]pyruvateintheperfusedheart,andthetypicalkineticprogressionof[1-13C]pyruvatemetabolites,isshowninFigure3.Followinginfusionof[1-13C]pyruvateintocontrolhearts,[1-13C]lactate(183.2ppm),H13CO23(160.9ppm),

and[1-13C]alanine(176.5ppm)wereclearlydetectablewithhighsignalcomparedwiththebaseline.Aresonancecorrespondingto13CO2wasalsovisible,with1stemporalresolution,atachemicalshiftof124.5ppm.Theinitialratesofproductionandthemaximumpeakareasforthe[1-13C]pyruvate-derivedmetabolites,inpre-ischaemic,ETZ,andreperfusedhearts,aregiveninTable1.

ThemaximumpeakareasofH13CO23,

13CO2,andtheirsumwerenotsignificantlydifferentfrombaselinewhen[1-13C]pyruvatewasinfusedintothemyocardiumuponreperfusion.However,theinitial

rateofH13CO23plus13CO2productionwas54%sloweruponreperfu-sion,comparedwiththepre-ischaemicmyocardium,asindicatedbythe

85Figure2ChangesinATP,PCr,Pi,andpHi,before,during,andafterischaemia.ATPlevelsgraduallydecreasedduringischaemia,andafterreperfusion,partiallyrecoveredtopre-ischaemialevels.PCrlevelsrapidlydecreasedattheonsetofischaemiaandrapidlyrecoveredtopre-ischaemiclevelsafterreperfusion.Pilevelsrapidlyincreasedattheonsetofischaemiaandrapidlydecreasedtopre-ischaemiclevelsafterreperfusion.pHigraduallydecreasedfrom7.07to6.49duringischaemiaandrapidlyrecoveredtopre-ischaemialevelsfollowingreperfusion.slopeofthereperfusionpeaksshowninFigure4.Additionally,thedecay

rateofhyperpolarized13CO2signalwas30%fasterinreperfusedheartsthaninpre-ischaemichearts(P,0.001),indicatingenhancedCO2washoutuponre-flowafterischaemia.

ETZhadnosignificanteffectontheinitialrateofH13CO23plus13CO2production,orthemaximumpeakareaofthesumofH13CO23and13CO2,comparedwithpre-ischaemichearts(Figure4).However,ETZincreasedthemaximum13CO2peakareabyfour-fold,whereasdecreasingthemaximumH13CO23peakareabytwo-fold(Table1,P,0.001).Additionally,thedecayrateofhyper-polarized13CO2signalwas19%fasterinreperfusedheartsthaninpre-ischaemichearts(P,0.05),possiblyindicatingenhancedCO2dif-fusionoutofmyocytesintheabsenceofCAactivity.

3.3MeasurementofpHiintheisolatedperfusedheart

Figure5AshowsthechangesinH13CO23and

13CO2,bothnormalizedtothemaximum[1-13C]pyruvatesignal,whichwereusedforthe

86

M.A.Schroederetal.

Figure3(A)Representativespectrumacquiredduringhyperpolarized[1-13C]pyruvateinfusionintotheisolatedperfusedheart.Fivesingle1s

spectraweresummedtoyieldthisspectrum,acquiredusinga308RFpulse.(B)Changesinthemetabolicproductsof[1-13C]pyruvateinpre-ischaemichearts(n¼6).

calculationofpR.Whenhyperpolarized[1-13C]pyruvatereachedthe

13isolatedheart,metabolicallygeneratedH13CO2CO2wereout3and

ofequilibriumfor󰀂5sbeforepR[Eq.(2)]reachedasteady-statevalueof7.12+0.10(Figure5B).Fullyrelaxed31Pmeasurements,acquiredinthepre-ischaemicheart,gaveapHiof7.07+0.02.ThepHimeasuredusing31PMRSandthe95%confidenceintervalareoverlaidonthe13CresultsinFigure5B.31PMRSconfirmedthatCAinhibitionwithETZhadnoeffectonsteady-statemyocardialpHi(pHiof7.02+0.03beforeETZtreat-mentand7.00+0.04afterETZtreatment).Perfusionwith[1-13C]pyruvateandETZgeneratedmore13CO2thanH13CO2313132(Figure5C)withnochangeintotalHCO3plusCO2.ThepR,cal-culatedfromtheH13CO23/CO2ratio,stabilizedwithin20stoasteady-statepRof6.21+0.13(Figure5D).Thus,inhibitionofCAactivityslowedCO2–HCO23conversion,asshownbythelengtheningoftheout-of-equilibriumperiod,butalsobythesteady-statepRwhichwas0.79pHunitsbelowthepHimeasuredusing31PMRS.Inreperfusedhearts,31PMRSrevealedthatpHirecoveredfromavalueof6.49+0.04attheendofischaemiato7.04+0.13,atarateof0.73pHunits/minduringthe45simmediatelyafterreflow(Figure6).Inheartsreperfusedwithhyperpolarized[1-13C]pyruvate,

31thepRfromtheH13CO2P3/CO2ratiowasthesameaspHifromMRSafter15sofreperfusion,whenaveragedinto30ssegmentsthatcorrespondedwithacquisitionof31Pspectra.After45and75s,both13Cand31PmeasurementsgavealmostidenticalpHimeasurements(13C:7.01+0.01at45sand6.98+0.02at75s;31P:7.04+0.13at45sand7.00+0.04at75s,Figure6).

3.4Mathematicalmodellingofexperimentalresults

Resultsofthemathematicalmodelof13CO2production,efflux,andhydrationtoH13CO23byCAaredepictedinFigure7.Figure7Ashowstheoutputofthemodelthatbest-fitstheexperimentaldatapresentedinFigure5.ConstantsKCO2(1026.15M),kf(0.14s21),andkr(kf/KCO2)wereobtainedfrompublishedvalues21andotherpar-ameterswereobtainedbyleast-squaresfitting:Ppyr(0.2s21),PCO2(0.2s21),r(0.006s21),a(1/33sforpyruvate,1/6sforCO2and

13HCO2CO2and3).Thebestsimulationofourexperimental

213HCO3results(Figure7A)indicatedthatCAactivity(g)enhancedtheconversionrateof13CO2intoHþþH13CO23by10-foldinpre-ischaemichearts.ThisvaluewasinlinewithaninvitrostudythatreportedCA-enhancedCO2hydrationbyfive-foldinisolatedmyocytes.21Figure7BshowsthevalueofpR[Eq.(2)]derivedfromthesimu-lationsinFigure7A.AtnormalCAactivity(g¼10),pRapproached6.8within17s,givingareasonableapproximationtotherealpHiof7.1.However,intheabsenceofCAactivity(g¼1),pRapproachedasignificantlylowerasymptoteof6.1within24s.

ApartfromCAactivity,anotherfactorthatcandisturbtheequili-2briumbetweenHþ,HCO23andCO2isHCO3transport.Figure7CandDillustratestheimplicationsofHCO23extrusionanduptake,respectively,onthesteady-statevalueofpR.Inthepresenceof+5mM/mintransmembraneHCO23flux,thevalueofpRwasnotgreatlyaltered,comparedwithamodelwithnonetHCO23transport.

MeasuringcardiacintracellularpHusinghyperpolarizedmetabolites

87

Table1Metabolitelevelsandkineticparametersfrom13CMRspectrainpre-ischaemia,reperfused,andETZ-perfusedisolatedhearts

[1-13C]LactatePre-ischaemia6+10.7+0.135+4H13CO23................................................................

Reperfusion31+3†4.4+0.4†22.3+0.2*

ETZ

Maximummetabolite/pyruvate(%)Initialproductionrate(%/s)Decay,t(s)

7+20.7+0.241+5

[1-13C]AlaninePre-ischaemia

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

Reperfusion

ETZ

3.5+0.20.28+0.0341+1

5.7+0.40.7+0.0143+7

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

4.2+0.30.43+0.0939+2

13................................................................

Pre-ischaemia

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

Pre-ischaemia0.60+0.060.06+0.02

48+2

CO2...................

Maximummetabolite/pyruvate(%)Initialproductionrate(%/s)Decay,t(s)

4.7+0.60.49+0.0643+4

Reperfusion

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

ETZ

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

Reperfusion0.70+0.06

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

ETZ

3.8+0.40.21+0.04*35+2

2.1+0.2*0.14+0.02*44+4

2.6+0.2†..............

0.053+0.00733+2†0.31+0.03†39+3*

Dataareexpressedmeans+SEM.Allmetabolitelevels,andinitialproductionrates,areexpressedasapercentageofmaximum[1-13C]pyruvatesignal.Significantdifferencefrompre-ischaemichearts:*P,0.05and†P,0.001.

0.1andwasthusatthelimitofdetectablesignal.Therefore,tocalcu-latepR,1sinvivospectrawereaveragedingroupsoftwotoyieldasetofspectrawith2stemporalresolutionandtheSNRimprovedto

1316.9+3.5and2.0+0.4forH13CO2CO2,respectively.Using3and

theaveragedspectra,pRreachedasteady-statevalueof7.20+0.03,asshowninFigure8B.

4.Discussion

4.1PDHfluxbeforeandafterischaemia

TostudytheCO2/HCO23equilibrium,PDHfluxmustbesufficienttogenerateMR-detectablelevelsof13CO2.Therefore,ourfirstaimwastodeterminetheeffectofischaemiaonpyruvateoxidation.OthershavestudiedPDHfluxuponreperfusionoftheischaemicmyocar-dium,withdiverseresultsdependingontheischaemicmodelandtheperfusionconditions.17,24,29–31KobayashiandNeely29observedthatpyruvateplusglucoseperfusionlargelymaintainedPDHactivityintheisolatedreperfusedheart,andinvivoPDHactivitywasmain-tainedfollowingreductionofcoronaryflowinswine.32However,intheisolatedratheartperfusedwithpyruvatealoneorpyruvateandfattyacids,ischaemiadecreasedPDHactivityandglucoseoxidationforseveralminutesfollowingreperfusion.17,30,31Here,10minoftotalglobalischaemiadecreasedtherateofpro-13ductionofH13CO2CO2from0.57+0.06to0.26+0.05%/s,3plus

indicatingadecreaseintheinitialrateofpyruvateoxidation,andthusinhibitionofPDHactivityinreperfusion.However,asignificant

13decreaseinthetotalH13CO23andCO2producedwasnotobserved,suggestingthatPDHfluxrecoveredtocontrollevelswithin30s.Mostimportantly,sufficient13CO2wasproducedatthestartofreperfusion

13toallowtheH13CO23/CO2ratiotobemeasured,which,underappropriateconditions,maybeusedtoestimatepHi.

Figure4Comparisonofthetimecoursesforthesumofthe

bicarbonateandcarbondioxidepeaks,normalizedtothemaximumvalueofpyruvatepeakarea.Themaximumpeakareadidnotchangefollowingeitherintervention,comparedwiththecontrol.Followingischaemia,theinitialslopeofthecurvewassignifi-cantlyreduced.

ItisnoteworthythathyperpolarizedH13CO23isonlyasmallfraction

2oftotalHCO3,andonlyaminorfractionoftransmembraneHCO2313effluxwouldbelabelledwithhyperpolarizedC.

FurthermodellingexploredtherelationshipbetweenpHiandthesteady-statepR(Figure7E)andthetimerequiredforequilibration(Figure7F),measuredasthetimetakenforpRtoapproachsteady-statepRwithin0.05U.AsCAactivity(g)wasincreased,steady-statepRapproachedthetruepHianddidsowithasmallertimedelay.ForlowvaluesofCA(g,10),pRsignificantlyunderestimatedpHi.More-over,forvaluesofCA,10,thetimetakenforpRtoattainsteadystatewas10–20s,asignificantfractionofthelife-timeofhyperpolarized13Ccompounds25andthetimeofpHirecoveryfollowingischaemia.

3.5MeasurementofpHiinvivo

Arepresentativespectrumof[1-13C]pyruvateinfusedinvivoisshowninFigure8A.TheH13CO23wasobservedwithasignal-to-noiseratio(SNR)of9.6+1.1,whereasthe13CO2peakhadanSNRof1.2+

4.2CO2/HCO23equilibriumasameasureofcardiacpHi13WeconvertedtheH13CO23/CO2ratiointheisolatedperfusedratheart,andintheinvivoratheart,intoavariable,pR,usingtheHenderson–Hasselbalchequation.Atsteady-state,pRintheisolatedperfusedratheartwas7.12+0.1,similar,withinthenoiseinherentin

88

M.A.Schroederetal.

Figure5(A)Thebicarbonateandcarbondioxide,bothnormalizedtomaximumpyruvatepeakarea,vs.timeincontrolhearts(n¼6).The

13point-by-pointratioofthesespecieswasusedtocalculatepR.(B)MeasurementofpRbasedonH13CO23/CO2incontrolheartscomparedwithmeasurementofpHiusing31PMRSinthesamegroupofhearts.(C)Thebicarbonateandcarbondioxide,bothnormalizedtomaximumpyruvate

13peakarea,inheartsperfusedwiththeCAinhibitorETZ(n¼5).(D)MeasurementofpRbasedonH13CO23/CO2inETZ-perfusedheartscomparedwithmeasurementofpHiusing31PMRSinthesamegroupofhearts.Thesteady-statepRapproached6.21,avaluewhichunderestimatedthetruepHioftheheartby0.80pHunits.

Figure6ComparisonofthepRmeasurementsmadeinreper-13fusedheartsusingtheH13CO23/CO2ratioandthepHimeasure-mentsmadeusing31PMRS.Thex-axisshowsthetimeafterreperfusionwithhyperpolarized[1-13C]pyruvate.The31Pmeasure-mentatt¼0equalsthepHifollowing10minofsimulatedischaemia.EachpRmeasurementwascalculatedastheaverageof30sofspectra,acquired15sbeforeandaftertheequivalent31Pmeasure-menttoallowfordirectcomparison.

eachmeasurement,tothepHiof7.07+0.02measuredusing31PMRS.Whenmeasuredinratheartsinvivo,pRwas7.20+0.03.ThesevaluesaresimilartothoseofMerrittetal.,17andwithpHimeasuredbyothersusing31PMRS.8–10,20,28Thus,hyperpolarized[1-13C]pyruvatecanbeusedtoobtainanaccurate,non-invasivemeasurementofcardiacpHiinvivoinhealthyhearts.17AsafirsttestofthesuitabilityofpRtomeasurepHi,weinhibitedcardiacCAactivityinperfusedrathearts,withoutalteringpHi.Wefoundasignificantdifferencebetweenthesteady-statepRof6.21andthepHiof7.01determinedusing31PMRS.LowpHi,suchasthatobservedduringmyocardialischaemia,inhibitsCAactivity.22,23Thus,

13theH13CO23/CO2ratiowouldnotbeagoodmeasureofpHiintheischaemic/reperfusedheartwithoutcorrectionforlowCAactivity.

13BymodellingourH13CO2CO2results,weidentifiedthe3and

conditionsinwhichpRwasnotavalidmeasureofpHi.Providedthatsufficient13CO2isgeneratedviaPDHflux,factorsthataltertherateofCO2production(Ppyr,r)willnotaltersteady-statepR.Likewise,CO2efflux(PCO2)doesnotaffectsteady-statepRwhentheheartisperfusedtotheextentthatextracellular13CO2iswashedawayrapidly.ChangestoCO2permeabilitywillalterCO2andHCO23levelsinparallelandwillnotaffectsteady-statepR.

AdiscrepancybetweenpRandpHoccurredwithchangesinthe

þrateofCO2hydrationandeventsrelatedtoHCO23andH.The

MeasuringcardiacintracellularpHusinghyperpolarizedmetabolites

89

Figure7Theresultsofmathematicallymodellingourexperimentaldata,acquiredfromcontrolheartsandheartsperfusedwithETZ.(A)Results

fromthemodelwhichbest-fitourexperimentaldatafromFigures4Aand5A.Experimentaldatawerebestreproducedwitha10-foldcatalyticactivityofCA.(B)ThemodelsimulationofpR,basedonthetimecoursesfrom(A).WithmoderatelevelsofCAactivity,asmaybeexpectedintheheart,the

13H13CO23/CO2ratioindicatesasteady-statepRthatcloselyapproximatesthephysiologicalvalues.However,withlowerCAactivity,themodelreproducedourexperimentalfindingofpHiunderestimation.(C)TherelationshipbetweenCAactivityandsteady-statepR,inthepresence(dashedgrey)andabsence(blacksolid)ofHCO23effluxor(D)influx.CAactivityhasasignificanteffectonthesizeofthepR–pHidiscrepancy,

2butHCO3transporthasamuchsmallerimpactonthediscordance.(E)Relationshipbetweensteady-statepRandpHi,simulatedfordifferentlevelsofCAactivity.(F)RelationshipbetweenequilibrationtimeandpHi,simulatedfordifferentlevelsofCAactivity.EquilibrationtimewasestimatedasthetimetakenforpRtoapproachsteady-statepRwithin0.05U.

rateofCO2hydrationdependsonCAactivity.CAactivityincardiacmyocytesismodest21and,furthermore,CAisinhibitedbylowpH22,23andbypharmacologicalmembrane-transportinhibitors.33iConsequently,CO2hydrationkineticshaveanimpactonthesuit-abilityofpRasameasureofpHi.TheimportanceofCAactivitywastestedusingamathematicalmodel(Figure7).AtlowCAactivity,pRattainedasteady-statethatcouldbeverydifferentfrompHi,andtheequilibrationtimecouldbeasignificantfractionofthelifetimeofhyperpolarized13Corofrecoveryfromischaemia-inducedacidosis.MembranetransportofHþandHCO23,beingdown-streamofCO2hydration,alsodisplacedpRawayfrompHi.BecauseofhighpHibuffer-ingcapacityinsidemyocytes,2transportofHCO23wouldhavearela-tivelygreatereffectonthestateoftheCO2/HCO23equilibriumthanþHtransport.ThemathematicalmodelwasusedtoinvestigateHCO23transportatthemodestrateof+5mM/min(Figure7Cand

D).TheerrorduetoHCO23transportwas,however,negligibleand

isthereforetooslowtosignificantlyaffectCO2/HCO23equilibrium.

4.3Limitationsofthestudy

TomeasurepHiusing[1-13C]pyruvate,itisessentialthatthemetabo-13licallygeneratedH13CO2CO2resonancescanbeaccurately3and

quantifiedabovethebaselinenoise.AtaphysiologicalpHof󰀂7,

13theH13CO2CO2resonance,3resonanceis10-foldlargerthanthe

13soCO2quantificationrequiresefficienthyperpolarization,highPDHflux,andcarefuldataacquisition.Anincreaseinachievablepolarization,fromthe󰀂30%observedhereto󰀂60%,hasrecentlybeenreported19andwillaid13CO2quantification.Also,strategicdataacquisitionafterthe13Cequilibrationperiod,overashorterdur-ation,andwithahigherexcitationflipanglemayfurtherincreasethe13CO2signal.

90

M.A.Schroederetal.

Figure8(A)Representativeinvivospectrumacquiredduringhyperpolarized[1-13C]pyruvateinfusionintolivingrathearts.Twosingle1sspectra

13weresummedtoyieldthisspectrum,acquiredusinga308RFpulse.(B)MeasurementofinvivopRbasedonH13CO23/CO2inlivingrathearts.

Asecondlimitationofthisstudyisthefactthatthehyperpolarizedlabelcannotdirectlydistinguishbetweenmetaboliteslocatedintheintra-andextracellularspaces.Wecanbecertainthatduetohighcardiacoxidativerates,whicharemorethananorderofmagnitudehigherthananyneighbouringtissue(i.e.liver,restingskeletalmuscle,adiposetissue,diaphragm,orblood),virtuallyallofthe

13detectedH13CO2CO2signalwasproducedwithinthemyo-3and

cardium.However,itispossiblethattraceamountsof13CO2mayhavediffusedoutofthemyocardiumandweresubsequentlyhydratedtoformH13CO23eitherspontaneously,byextracellularcardiacCA,orinvivo,byCAinredcells.However,webelievethatthecontri-13butionoftheextracellularH13CO2CO2signaltoourpHi3and

measurementwassmallbecause:(i)invivospectroscopic13Cimagesacquiredoftheheart34haveindicatedthatH13CO23iscon-finedtothemyocardium,aregiondominatedbytheintracellularspace;(ii)intheperfusedheart,highcoronaryflowrates(󰀂20mL/min)wouldhaverapidlyremovedhyperpolarizedmetabolites,andinvivoassociationof13CO2withhaemoglobinwouldhavecausedrapiddecayofhyperpolarizedMRsignal;and(iii)thecloseagreementbetweenpRmeasuredwith13CandpHimeasuredwith31Pintheperfusedheartindicatedminimalcontaminationfromextracellular

13H13CO23andCO2,asthesespecieswouldhaveequilibratedaccord-ingthepHoof7.4.

4.4Significanceofthiswork

Currently,thenon-invasivemeasurementofcardiacpHiinhumansisimpossible,becauseblood2,3-DPGsignaloverliesthemyocardialPisignal.6,9,10Here,wehaveshownthatinthepresenceofendogenous

13CAactivity,theH13CO23/CO2ratioaccuratelymeasuredpHiintheisolatedperfusedheart.Further,wehavedemonstratedthatfollowing

infusionofhyperpolarized[1-13C]pyruvateintohealthyratsinvivo,the

13MRsignalcorrespondingtoH13CO2CO2couldbothbequan-3and

tified,andthattheirratioindicatedapHivalueof7.20,whichisinlinewithinvasivemeasurements.9,10Consequently,itseemsthatmeta-13bolicallygeneratedH13CO2CO2mayofferthefirsttechnique3and

forthenon-invasivemeasurementofpHiinnormalhearts,andindis-easedheartswithnormalorelevatedCAactivity.35MeasuringinvivopHialsoimpliesthatotheranalysesofmyocardialenergeticsmaybeperformedinvivo,includingcalculationoffreeADPconcentrationsandthefreeenergyavailablefromthehydrolysisofATP,DGATP.5,10Futureworkwillinvolvecorrelatinginvivomeasurementsofthe

13H13CO23/CO2ratiowithpHimeasurementsmadeusinginvasive,blood-removedoropen-chesttechniques.9,10Sinceacidosisisacharacteristicfeatureofischaemia,assessmentofischaemicheartdiseaseinhumansisanotherpotentiallyusefulappli-13cationofanon-invasivepHicalculationusingtheH13CO23/CO2ratio.WeobservedexcellentagreementbetweenpHimeasuredusing31PMRSandpRmeasuredusing13CMRSinthereperfusedmyocardiumwhenpHiwas!6.74(Figure6).However,multiplefactorsshiftpRrelativetopHi,includinginhibitionofCAactivityatlowpH22,23andstimulationofmembranetransportduringreperfu-isionfollowingischaemia.2,36Pharmaceuticalagents,suchascariporide,inhibitmembraneiontransportandhavebeenusedclinicallytoreduceischaemia–reperfusioninjury,37butalsoblockCA.33There-13fore,theuseoftheH13CO23/CO2ratiotomeasurepHimaynotbevalidinischaemic,acidichearts,andinpatientswithischaemicheartdiseasewhousedrugsthatinhibitmembraneiontransport.Additionally,itisreasonabletoexpectthatintracellularCAexpressionandactivitymaybeeitherreducedorincreasedinotherformsofcardiomyopathy.35CorrectionofpRtopHiwillrequire

MeasuringcardiacintracellularpHusinghyperpolarizedmetabolitesfullcharacterizationofCAactivityineachpathophysiologicalstate,andmathematicaldeconvolutionofthe13CequilibrationperiodfromthetruemeasuredpHichanges.EventualtranslationoftheH13CO23/13CO2ratiotomeasurepHiintheclinicwillrequireconsiderabletechnologicaladvances,intermsofimprovedmethodsandhardwareforacquisitionof13Cimages,andaccesstoaffordablehyperpolarizationtoolsand13C-labelledcom-pounds.InordertoidentifyfocalregionsofischaemiausingpHimeasurementsfromhyperpolarized[1-13C]pyruvate,forexample,

three-dimensionalimagesofH13CO23and

13CO2withrelativelyhighspatialresolutionacrosstheareaatriskwillberequired.Thefeasibilityofacquiringsuchdataacrossthemyocardiumoflargeanimals,andthereforepatients,hasbeendemonstrated.34Further,althoughtheeventualcostofclinicalapplicationofthehyperpolarized13CMRtechnologyisnotclear,itdoesnotappearsettobeprohibi-tive.Clinicalpolarizerscouldbeoperatedasstandalonesystems,placedwithinexistingclinicalMRfacilitiesandinterfacedtoexistingMRscanners.Further,thecostof[1-13C]pyruvicacid,asusedhere,isnotexcessiveandwouldbeinlinewithcontrastagentsusedinotherimagingmodalities,suchaspositronemissiontomography.Insummary,wehavedemonstratedintheperfusedheartthattheH13CO23/13CO2ratiooffersanaccuratemethodtomeasurecardiacpHiinheartswithnormalorelevatedCAactivity.Further,thetech-niqueappearssettobecomethefirstclinicallyrelevantmeasureofinvivocardiacpHi,althoughfutureworkiswarrantedtocharacterizeCA

activityandtheresponseoftheH13CO23/13CO2ratioinischaemiaandothercardiomyopathies,andtoimprovethesensitivityand

spatialresolutionofH13CO23and

13CO2detection.Supplementarymaterial

SupplementarymaterialisavailableatCardiovascularResearchonline.

Acknowledgements

TheauthorswouldliketothankProf.RichardVaughan-Jones,Prof.KevinBrindle,andDrJan-HenrikArdenkjær-Larsenforhelpfulsuggestions.Conflictofinterest:ThisworkreceivedresearchsupportfromGE-Healthcare.

Funding

M.A.S.isfundedbytheNewtonAbrahamScholarshipFoundation,NIHgrantno.1-F31-EB006692-01A1andtheWellcomeTrust.P.S.isfundedbytheMedicalResearchCouncilandtheRoyalSociety.F.A.G.isfundedbyCancerResearchUKandtheNationalInstituteofHealthResearchCambridgeBiomedicalResearchCentre.ThisworkwasfundedbygrantsfromtheMedicalResearchCouncil(MRCGrantG0601490)andtheBritishHeartFoundation(BHFGrantPG/07/070/23365),andbyGE-Healthcare.FundingtopaytheOpenAccesspublicationchargesforthisarticlewasprovidedbytheWellcomeTrust.

References

1.GarlickPB,RaddaGK,SeeleyPJ.Studiesofacidosisintheischaemicheartbyphos-phorusnuclearmagneticresonance.BiochemJ1979;184:547–554.

2.Vaughan-JonesRD,SpitzerKW,SwietachP.IntracellularpHregulationinheart.JMolCellCardiol2009;46:318–331.

3.OpieLH.Myocardialischemia–metabolicpathwaysandimplicationsofincreasedgly-colysis.CardiovascDrugsTher1990;4(Suppl.4):777–790.

4.BingOH,BrooksWW,MesserJV.Heartmuscleviabilityfollowinghypoxia:protec-tiveeffectofacidosis.Science1973;180:1297–1298.

5.IngwallJ.ATPandtheHeart.Boston:KluwerAcademicPublishers;2002.

6.FrohlichO,WallertMA.MethodsofmeasuringintracellularpHintheheart.Cardio-vascRes1995;29:194–202.

917.HoultDI,BusbySJ,GadianDG,RaddaGK,RichardsRE,SeeleyPJ.Observationoftissuemetabolitesusing31Pnuclearmagneticresonance.Nature1974;252:285–287.8.BaileyIA,WilliamsSR,RaddaGK,GadianDG.Activityofphosphorylaseintotalglobalischaemiaintheratheart.Aphosphorus-31nuclear-magnetic-resonancestudy.BiochemJ1981;196:171–178.9.KatzLA,SwainJA,PortmanMA,BalabanRS.IntracellularpHandinorganicphosphatecontentofheartinvivo:a31P-NMRstudy.AmJPhysiol1988;255:H189–H196.

10.BrindleKM,RajagopalanB,WilliamsDS,DetreJA,SimplaceanuE,HoCetal.P-31NMR

measurementsofmyocardialpHinvivo.BiochemBiophysResCommun1988;151:70–77.11.GallagherFA,KettunenMI,DaySE,HuDE,Ardenkjaer-LarsenJH,ZandtRetal.Mag-neticresonanceimagingofpHinvivousinghyperpolarized13C-labelledbicarbonate.Nature2008;453:940–943.

12.LeemCH,Lagadic-GossmannD,Vaughan-JonesRD.Characterizationofintracellular

pHregulationintheguinea-pigventricularmyocyte.JPhysiol1999;517:159–180.13.SwietachP,Vaughan-JonesRD,HarrisAL.RegulationoftumorpHandtheroleof

carbonicanhydrase9.CancerMetastasisRev2007;26:299–310.

14.HoffmanDW,HenkensRW.Theratesoffastreactionsofcarbondioxideandbicar-bonateinhumanerythrocytesmeasuredbycarbon-13NMR.BiochemBiophysResCommun1987;143:67–73.

15.MadshusIH.RegulationofintracellularpHineukaryoticcells.BiochemJ1988;250:1–8.16.MerrittME,HarrisonC,StoreyC,JeffreyFM,SherryAD,MalloyCR.Hyperpolarized

13Callowsadirectmeasureoffluxthroughasingleenzyme-catalyzedstepbyNMR.ProcNatlAcadSciUSA2007;104:19773–19777.

17.MerrittME,HarrisonC,StoreyC,SherryAD,MalloyCR.Inhibitionofcarbohydrate

oxidationduringthefirstminuteofreperfusionafterbriefischemia:NMRdetectionofhyperpolarized13CO2andH13CO3.MagnResonMed2008;60:1029–1036.

18.SchroederM,CochlinL,HeatherL,ClarkeK,RaddaG,TylerD.Invivoassessmentof

pyruvatedehydrogenasefluxintheheartusinghyperpolarizedcarbon-13magneticresonance.ProcNatlAcadSciUSA2008;105:12051–12056.

19.JohannessonH,MachollS,Ardenkjaer-LarsenJH.Dynamicnuclearpolarizationof

[1-13C]pyruvicacidat4.6tesla.JMagnReson2009;197:167–175.

20.VandenbergJI,CarterND,BethellHW,NogradiA,RidderstraleY,MetcalfeJCetal.

CarbonicanhydraseandcardiacpHregulation.AmJPhysiol1996;271:C1838–C1846.21.LeemCH,Vaughan-JonesRD.Out-of-equilibriumpHtransientsintheguinea-pigven-tricularmyocyte.JPhysiol1998;509:471–485.

22.KernohanJC.ThepH-activitycurveofbovinecarbonicanhydraseanditsrelationship

totheinhibitionoftheenzymebyanions.BiochimBiophysActa1965;96:304–317.23.KhalifahRG.Thecarbondioxidehydrationactivityofcarbonicanhydrase.I.Stop-flow

kineticstudiesonthenativehumanisoenzymesBandC.JBiolChem1971;246:2561–2573.

24.SchroederMA,AthertonHJ,BallDR,ColeMA,HeatherLC,GriffinJLetal.Real-time

assessmentofKrebscyclemetabolismusinghyperpolarized13Cmagneticresonancespectroscopy.FASEBJ2009;23:2529–2538.

25.Ardenkjaer-LarsenJH,FridlundB,GramA,HanssonG,HanssonL,LercheMHetal.

Increaseinsignal-to-noiseratioof.10,000timesinliquid-stateNMR.ProcNatlAcadSciUSA2003;100:10158–10163.

26.NaressiA,CouturierC,CastangI,deBeerR,Graveron-DemillyD.Java-basedgraphi-caluserinterfaceforMRUI,asoftwarepackageforquantitationofinvivo/medicalmag-neticresonancespectroscopysignals.ComputBiolMed2001;31:269–286.

27.MurrayAJ,LygateCA,ColeMA,CarrCA,RaddaGK,NeubauerSetal.Insulinresist-ance,abnormalenergymetabolismandincreasedischemicdamageinthechronicallyinfarctedratheart.CardiovascRes2006;71:149–157.

28.CrossHR,ClarkeK,OpieLH,RaddaGK.Islactate-inducedmyocardialischaemic

injurymediatedbydecreasedpHorincreasedintracellularlactate?JMolCellCardiol1995;27:1369–1381.

29.KobayashiK,NeelyJR.Effectsofischemiaandreperfusiononpyruvatedehydrogen-aseactivityinisolatedrathearts.JMolCellCardiol1983;15:359–367.

30.LewandowskiED,WhiteLT.Pyruvatedehydrogenaseinfluencespostischemicheart

function.Circulation1995;91:2071–2079.

31.LopaschukGD,SpaffordMA,DaviesNJ,WallSR.Glucoseandpalmitateoxidationin

isolatedworkingratheartsreperfusedafteraperiodoftransientglobalischemia.CircRes1990;66:546–553.

32.StanleyWC,HernandezLA,SpiresD,BringasJ,WallaceS,McCormackJG.Pyruvate

dehydrogenaseactivityandmalonylCoAlevelsinnormalandischemicswinemyocar-dium:effectsofdichloroacetate.JMolCellCardiol1996;28:905–914.

33.VillafuerteFC,SwietachP,Vaughan-JonesRD.CommoninhibitorsofmembraneHþ-transportalsoinhibitcarbonicanhydrase.FASEBJ2007;21:A1284–A1284.

34.GolmanK,PeterssonJS,MagnussonP,JohanssonE,AkesonP,ChaiCMetal.Cardiac

metabolismmeasurednoninvasivelybyhyperpolarized13CMRI.MagnResonMed2008;59:1005–1013.

35.AlvarezBV,JohnsonDE,SowahD,SolimanD,LightPE,XiaYetal.Carbonicanhy-draseinhibitionpreventsandrevertscardiomyocytehypertrophy.JPhysiol2007;579:127–145.

36.VandenbergJI,MetcalfeJC,GraceAA.MechanismsofpHirecoveryafterglobalische-miaintheperfusedheart.CircRes1993;72:993–1003.

37.ScholzW,AlbusU,CounillonL,GogeleinH,LangHJ,LinzWetal.Protectiveeffects

ofHOE642,aselectivesodium-hydrogenexchangesubtype1inhibitor,oncardiacischaemiaandreperfusion.CardiovascRes1995;29:260–268.