SEPARATEGENESDETERMININGTHESTRUCTUREANDINTRACELLULARLOCATIONOFHEPATICG - PDF document

1 SEPARATEGENESDETERMININGTHESTRUCTUREANDI
SEPARATEGENESDETERMININGTHESTRUCTUREANDINTRACELLULARLOCATIONOFHEPATICGLUCURONIDASE*BYROGERGANSCHOWtANDKENNETHPAIGENDEPARTMENTOFEXPERIMENTALBIOLOGY,ROSWELLPARKMEMORIALINSTITUTE,BUFFALO,NEWYORKCommunicatedbyBarbaraMcClintock,June20,1967Amajorfactorinfluencingthecontributionofanenzymetotheeconomyofthecellinwhichitresidesisthesiteatwhichtheenzymeislocatedwithinthecell.Animportantquestiontheniswhetherthepredilectionofenzymesforcertainintracellularsitesderivesonlyfromthechemicalstructureoftheenzymeorwhetheradditionalandspecificenzymelocalizingfactorsoccur.Inthisreportwepresentevidencethatsuchaseparateanddistinctfactorexistsfortheenzyme,-glucuroni-dase.Thestructureofthisenzymeininbredstrainsofthehousemouse,Musmusculus,isspecifiedbyasinglegeneonlinkagegroupXVII.,2Theenzymeresideswithinboththelysosomesandthemembranesoftheergastoplasminlivercells,andenzymemoleculesresidentatthetwositesarephysicallyindistinguishableandcodedforbythesamegene.3Amutationofthisgene,originallydiscoveredasamarkedreductionof,3-glucuronidaseactivityinalltissues,4alsoresultsinachangeinthethermalstabilityoftheenzyme.3Mutationofthestructuralgenesignificantlyalterstheproportionoftheactivityatthetwosites.3Notsurprisinglythen,thestructureofaproteinapparentlycanaffectitsownintracellulardistribution.Theresultsofthepresentstudydescribeanewgeneinthehousemouse,distinctfromtheglucuronidasestructuralgene,whichdetermineswhetherornotglucuroni-dasecanbeintegratedintotheergastoplasmicmembrane,thusdemonstratingtheexistenceofadditionalenzymelocalizingfactors.Apreliminaryreportofseveraloftheseexperimentshasappearedpreviously.'MaterialsandMethods.-Animals:DBA/2,C3H/HeHa,andYBRmice90-120daysofagewereallgenerouslysuppliedfromthebreedingcolonyofDr.T.S.Hauschka.Preparationoftissuehomogenates:Homogenates(5%)werepreparedincold0.25Msucrose,bufferedatpH7.4with0.02Mimidazole-HCl,usingaPotterElvehjem-typehomogenizer.Osmoticshock:Homogenateswerecentrifugedat105,000gfor30mintoseparateparticulatefromsolubleenzymes.Thesedimentwasresuspendedinhypotonicbuffer(0.02Mimidazole-HCl,pH7.4)inordertosolubilizelysosomalenzymes.Solubilizedenzymewasthenseparatedfromenzymeresistanttoosmoticshockbycentrifugingtheresuspendedparticlesat105,000gfor30min.Enzymeactivitiesofthefirstsupernatantandthefinalsupernatantandsedimentweredetermined.Differentialcentrifugation:ThescheduleofdeDuveetal.6wasfollowedusingaSpincono.40rotor.Fourparticulatefractionswereseparatedinorderofdecreasingsize:nuclear,mitochon-drial,lysosomal,andmicrosomal,plusthefinalsupernatant.Forsomeexperimentsacombinedcytoplasmicparticlefra

2 ctionwaspreparedbycentrifugingat105,000g
ctionwaspreparedbycentrifugingat105,000gfor30minafterremovalofthenuclearfraction.Zonecentrifugation:Threemillilitersoftheresuspendedcytoplasmicparticlefractionwerelayedoveranonlinear,31-mlgradientofsucrose(seeFig.1),ranginginconcentrationfrom0.4Mto2.1MandbufferedatpH7.5with0.02Mglycylglycine-NaOH.ThegradientwaspreparedaccordingtothesuggestionsofBockandLing.7Electrophoresisofmicrosomalprotein:Sedimentedmicrosomeswereresuspendedin0.02Mimidazole-HClbuffer,pH7.4,andthesuspensionwasmade1%indeoxycholate.Thismixturewasallowedtostandatroomtemperaturefor15minandthencentrifugedat105,000gfor60938 GENETICS:GANSCHOWANDPAIGENmin.Deoxycholatewasremovedfromthesupernatantfluidbydialysisagainsttwo1000-mlchangesoftheimidazole-HClbuffer.Thedialyzedextractofmicrosomalproteinwasthensub-jectedtoelectrophoresisinpolyacrylamide-gelcolumnsaccordingtothemethodofDavis8forserumprotein.TheproteinpreparationwasseparatedatpH9.5usingaconstantcurrentof5mapergelduringthe30-minrun.Thegelswerethenstainedforproteinwithanilineblue-black,andunboundstainwasremovedelectrophoretically.Thestainedgelswerestoredin7%aceticacidandphotographed.Enzymeactivitymeasurements:Glucuronidase(I.U.B.3.2.1.31)wasassayedat560CbyamodificationoftheprocedureofNimmo-Smith9whichemploysthesubstratep-nitrophenylglucuronide(CycloChemicalCo.).Thefinalreactionmixturecontainedenzyme,0.001Mp-nitrophenylglucuronide,0.1MacetatebufferpH4.6,and0.1%ofthesurfactantTritonX-100(RohmandHaasCo.).Thelatterwasaddedtoensurecompleteactivationoflatentlysosomalenzymes.'0Glucose-6-phosphatase(I.U.B.3.1.3.9)activitywasdeterminedat30'Cbymeasuringtheamountofphosphorus"liberatedbytheenzymatichydrolysisofglucose-6-phosphate.Thefinalreactionmixturecontainedenzyme,0.04Mglucose-6-phosphate(SigmaChemicalCo.),and5X10-4MEDTA,andwasbufferedwith3.5X10-'Mhistidine(pH6.5).Furtherdetails:Amoreextensivedescriptionoftheseexperimentalproceduresiscontainedinthereportofarelatedstudy."2Results.-Duringasearchforgeneticvariantsof,B-glucuronidase,theYBRstrainofinbredmicewasfoundtohavelittle,ifany,glucuronidaseactivityintheergastoplasmoflivercells."2ThisstrainhasbeenmaintainedinthepastbecauseitcarriestheAl(yellow-lethal)mutation,thedominantalleleoftheagoutigene.Ayislethalinthehomozygousconditionandcausesyellowcoatcolorandobesitywhenheterozygous.Thealleleismaintainedbycrossingtheyellowheterozygotes(A"a)withnonagouti(aa)animals.Thegeneticfactoraffectingglucuronidasewasfoundtobeindependentofthestateoftheagoutilocus,sinceallanimalsofthisstrainwereaffected.Bypreference,onlythenonagoutisegregantsoftheYBRstrainwereusedinthisstudy.Th

3 eupsetintheintracellulardistributionofgl
eupsetintheintracellulardistributionofglucuronidasewasfirstdetectedbyfindingthatalloftheparticulateglucuronidaseofaYBRliverhomogenatewasreleasedintosolutionafterosmoticshock.Incontrast,approximately40percentoftheparticulateenzymeofmostotherstrainsremainsparticulateafterosmoticshock.12Theosmoticshocktechniqueresultsinthesolubilizationoflysosomalenzyme,butnotmicrosomalenzyme.5Sincemicrosomesarisebyfragmentationoftheergastoplasmatthetimeofhomogenization,thisresultsuggestsadeficiencyofergastoplasmicenzymeinlivercellsofYBRmice.Directevidenceforthiswasobtainedusingsucrosegradientfractionation.Whenliverparticlesarecentrifugedthroughanappropriatenonlinearsucrosegradient,twobandsofparticlesrapidlyseparate.Lysosomesandmicrosomeshavebeenidentifiedinthelowerandupperzones,respectively,onthebasisofthedistributionsofmarkerenzymes,therelativecontentofthesebandsamongtheparticulatefractionsobtainedafterdifferentialcentrifugation,andtheirrespectivesensitivitiestoosmoticshock.Theseexperimentsarereportedelsewhere.'2Usingliverparticlesfromnormalmice,theresultingdistributionofglucuronidaseinthegradientisbimodal(Fig.1).If,inYBRmice,microsomalglucuronidaseislacking,butlysosomalenzymeisretained,weshouldexpecttofindonlyasingle,rapidlysedimentingpeakwhenYBRparticlesaresedimentedthroughasimilargradient.Figure2showsthatglucuronidaseactivitywasindeedlackingintheregionofthemicrosomesandthatonlyasinglepeakcorrespondingtothelysosomes939VOL.58,1967 GENETICS:GANSCHOWANDPAIGEN30�DBAcLow~~~~~~~~~~Wj0oSUCROSEIO00:H--LCrs0~~~~~~~~~~~~~~05(layeredoveranonlinearsucroseconcentrationgradientandsedimentedat24,000gfor30mlin.Fractions(3mleach)werecollectedandassayedforglucuronidase(o-k)andglucose-6-phosphatase(.---e)activities.Activityineachfractionisexpressedintermsofpercentofrecoveredactivityasafunc-tionofvolumepreviouslyremovedfrombottomofgradient.Sucroseconcentrationisdesignatedbyopentriangles(a-A).Enzymerecoveries:glucuronidase,14.3/15.8activityunits;glucose-6phosphatase,111/180activityunits.YBR:aaa.o:aFIG.2.-DistributionofenzymesF\G.:-DistributionofenzymeafterdensitygradientcentrrfugauonugationofcytoplasmicofcytoplasmicparticlesfromYBRlayeredoveanonlliver.Proceduresandsymbolsaresedimenteat24,000:for30min.FracidenticaltothoseofFig.1.Enzymeco--ectedandassayedforglucuronidrecoveries:glucuronidase,4.2/4.8ac-Enyereoeis:guuond,1431.tivityunits;glucoseausphosphatase,gs,^1,/0activityunits124/160activityunits.ML.FROMBOTTOM940PROC.N.A.S. GENETICS:GANSCHOWANDPAIGENTABLE1INTRACELLULARDISTRIBUTIONOFGLUCURONIDASEINDBAANDYBRMOUSELIVERPercentHomog.Nuc.Mit.Lys.M

4 ic.Super.recoveryDryweight(%total)DBA(36
ic.Super.recoveryDryweight(%total)DBA(360mgtotal)14226203799YBR(374mgtotal)14187203796Totalactivity(X103)DBA63.69.0819.711.713.49.1099YBR23.81.868.684.930.786.0594SpecificactivityDBA1776488951950669246YBR63.713348370539.2164SpecificactivityratioYBR/DBA0.360.200.540.360.060.67Totalactivitiesareexpressedasmicromoles-nitrophenolformedperhourat560Cpergramoftissue.Spe-cificactivitiesareexpressedasmicromolesp-nitrophenolformedperhourat560Cpermgdryweight.RelativespecificactivityistheactivitypermgdryweightofanyYBRfractiondividedbythatofthesameDBAfrac-tion.DryweightreferstotheTCA-insolublematerialineachfraction.waspresent.Thedistributionofthemicrosomalmarker,glucose-6-phosphatase,wasnormal.Theglucuronidaseactivityatthetopofthegradientissolubleandpresumablyarosefromdamagedlysosomes.IndependentconfirmationofthisconclusionwasobtainedwhenliverhomogenatesofYBRandwild-typeDBA/2werefractionatedbydifferentialcentrifugation,andtheglucuronidaseactivityofeachparticulatefractionwasdetermined.YBRactivitylevelsinnuclei,mitochondria,lysosomes,andsupernatantparalleledthoseofDBA,whereasthemicrosomalfractionofYBRhadasubstantiallyloweractivity(Table1).Thus,threedifferentcriteria(osmoticshock,gradientcentrifugation,andbulkfractionation)indicatethatlittle,ifany,glucuronidaseisincorporatedintotheergastoplasmofYBRlivercells.Moreover,thedatareportedinTable1suggestthattheenzymemissingfromthemicrosomeshasnotbeendisplacedtoanotherfraction.Thelackofglucuronidaseintheergastoplasmdoesnotreflectamajorupsetintheaccumulationofotherproteinsatthisintracellularsite,andseverallinesofevidenceindicatethattheanomalyisspecifictoglucuronidase.Firstofall,themicrosomalfractionobtainedbydifferentialcentrifugationofYBRliverhomoge-natescontainednormalamountsoftrichloraceticacid-precipitablematerial(Table1).Furthermore,theactivityofglucose-6-phosphatase,anothermicrosomalen-zyme,wasnormalinYBRliverhomogenatesandwasdistributedlikethatofwild-typeDBA/2inasucrosedensitygradient(Fig.2).Finally,diskelectrophoresisofproteinextractedfromYBRmicrosomesshowedapatternindistinguishablefromthatobtainedwithC3HmiceandfromF1progenyofthecrossbetweenthesetwostrains(Fig.3).AllproteinbandsobservedintheC3HelectropherogramappeartobepresentinsimilaramountsinthatofYBRandofF1.Theconclusionthattheanomalyisspecifictoglucuronidaseanddoesnotinvolvethebulkofthemicrosomalproteinraisesthequestionofwhetherthisenzymeisnolongercapableofbeingincorporatedintothemembranesoftheergastoplasmbe-VOL..58,1967941 GENETICS:GANSCHOWANDPAIGENC3HFYBRA=-_._~~~~ORIGIN__0F~~~~~~0.7-\|A~~~~~~.g~ggYgC0-j0.2+~inin~~11SERUM3

5 0.56-ALBUMINMIN.AT70.900FIG.3.-Diskelect
0.56-ALBUMINMIN.AT70.900FIG.3.-Diskelectrophoresisoflivermicro-FIG.4.-Heatinactivationofcellfractions.somalprotein.Approximately19proteinSurvivalofC3H(gcgc)andYBR(gygy)formsofbandscanbevisuallydistinguishedintheglucuronidaseactivityasafunctionofthedura-gels.tionofexposuretoheatingat70.9CC.Theshadedarearepresentstherangeofenzymeac-tivityfromeachofthecellularfractionsafterinactivation.ThefractionswerederivedfromthetheF,(gbgc)bythedifferentialcentrifugationschemeofdeDuveetal.tcauseofachangeinitsstructure.Toanswerthisquestion,theYBRstrainwascrossedwiththeC3H/HeHastrainwhichpossessesaformofglucuronidasedis-tinguishablefromthatoftheYBRbyitsrapidinactivationat71sC.312BothparentalformsoftheenzymecanbeidentifiedintheFrprogenyofsuchacross.IfYBRenzymecannotbeinsertedintotheergastoplasm,thenonlyheat-labileglucuronidaseactivityshouldbefoundinthemicrosomalfractionofF1progeny.Figure4showstheheatinactivationcurvesofeachparentalformofliverenzyme,aswellasthatoftheenzymeactivitypresentineachofthesubcellularfractionsobtainedbythedifferentialcentrifugationofF1homogenates.ThedenaturationcurvesforallF1fractionsfellwithinanarrowrange,indicatingthattherelativeproportionsoftheYBRandC3Htypesofenzymemoleculesweresimilarinallfractions.Furthermore,asthepositionoftheheat-denaturationcurveshows,thepreponderantformoftheenzymepresentinallfractionswasthatoftheYBRparent.Thus,theenzymepresentinthemicrosomesoftheF1progenyisprimarilyoftheYBRtype,implyingthatafunctionalalleleofsomegenenecessaryfortheinsertionoftheYBRformoftheenzymeintotheergastoplasmhasbeenrestoredbycrossingwiththeC3Hstrain.InordertodecidewhetherthegeneticfactormissinginYBRwastheC3Hglucuronidasestructuralalleleoranothergeneticelement,theprogenyoftheback-crossdiagrammedinFigure5wereexamined.Iftheessentialfactorweretheglucuronidaseallele,onlytwoclassesofbackcrossanimalswouldbeexpected:enzymeheterozygoteswithmicrosomalenzymeandhomozygoteswithnomicro-942PROC.N.A.S. GENETICS:GANSCHOWANDPAIGENg9gCm9c+)...T.ggY(mic)FIG.5.-Backcrossscheme.TheIsymbolgcreferstothemutantglucu-gCgY(mic(igygY(ronidasestructuralallelecharacteris--9~~~~~~~~ticofC3HmiceandgVtothatofYBRanimals.ThepresenceorabsenceofmicrosomalenzymeineachgeneticclassCgyygYisindicatedinparentheses.somalenzyme.However,ifthefactorweredistinctfromthestructuralgene,thentherecombinantclassesshouldalsooccur.Inordertodeterminewhichalleleswerepresentattheenzymestructurallocusinbackcrossprogeny,theinactivationrateofspleenenzyme,ratherthanliver,wasdetermined,sincetheC3Hformoftheenzymeispresentinconsiderablygreateramountsinspleenthaninanyotherorga

6 n."3Thepresenceorabsenceofmicrosomalenzy
n."3Thepresenceorabsenceofmicrosomalenzymeinliverwasdeterminedbytheosmoticshocktechniqueratherthanbydifferentialcentrifugationofcellparticles.Becauseofthecross-contaminationoffractions,differentialcentrifuga-tiondoesnotgiveanaccuratemeasureoftheabsoluteamountofenzymepresentateachsite.Thetwotechniques,inactivationofspleenenzymeandliver-particleosmoticshock,werecombinedtoyieldtheresultsshowninFigure6.Thedashedlinesrepresenttheobservedratesofdenaturationofthespleenglucuronidasefrommiceofknowngenotypes.Theshadedareasindicatetherangeofheatdenaturationratesofspleenenzymeinbackcrossanimals.Twodistinctclassesofinactivationcurvesarepresentcorrespondingtothetwoexpectedenzymeclasses.Theintra-cellulardistributionofliverglucuronidaseactivity,tabulatedontherightsideofFigure6,showsthateachenzymeclassincludedindividualswithmicrosomalglucuronidaseandindividualslackingmicrosomalglucuronidase.TheexistenceofthegYgY(Mic+)classofbackcrossprogenyshowsthatYBR-enzymehomozygotesMIN.AT71.0C2w15304560-JUSA:.LIVERGLUCURONIDASEACTIVITY7*gygLys.Mic.Sol.Recov.(%)2'.MICE23.7.416.810040.5.:_)14.60.515.69412.MIC+28.517.315.0101_MI~~~~~~~C-JT10.30.57.21020Q03%86.80.56.699ggMIC+J17.210.29.9102Li_0(5)13.59.612.698FIG.6.-Recognitionofbackcrossclasses.Spleenenzymesurvivalisrep-resentedasafunctionofthedurationofexposuretotheenzymetoheatat71.00C.Dashedlinesrepresenttheobservedratesofheatdenaturationofspleenglucuronidaseofmicewithknowngenotypes.Theshadedareasrepre-senttherangesofheat-denaturationratesseenamongbackcrossanimals.Glucuronidaseactivitiesofthelysosomal,microsomal,andsolubleliverfractionsderivedfrom1gmofliverofcorrespondingbackerossanimalsareindicatedontheright.Thenumberofanimalsexaminedineachgroupisindicatedinparentheses.VOL..58,1967943 GENETICS:GANSCHOWANDPAIGENcanacquiretheabilitytoinsertglucuronidaseintomicrosomesaswellasintolysosomes.TheexistenceofthegYgC(Mic-)classindicatesthatthisfactorisequallyrequiredfortheincorporationintotheergastoplasmofeitherstructuralformoftheenzyme.Thus,thegeneticfactorrequiredforinsertionofglucuronidaseintotheergasto-plasmassortsindependentlyoftheglucuronidasestructuralgene.ThesymbolEisproposedtorepresentthisfactor.Thewild-typealleleissymbolizedasE+,whilethemutantallele,whichresultsinanomissionofglucuronidasefromtheergasto-plasm,issymbolizedaseg'.SincetheEgeneassortsindependentlyofthepreviouslymappedglucuronidasestructuralgene,linkagegroupXVIIiseliminatedasaprobablelocationsite.LinkagegroupsVandVIIIarealsoexcludedsincetheagoutiandbrowngenes,segregatinginthebackerossprogeny,assortedindependentlyofE.Furthermore,an

7 examinationofthesexratiosandpedigreesoft
examinationofthesexratiosandpedigreesofthebackcrossanimalsindicatedthatthegeneisneithersex-linkednormaternallyinherited.InadditiontothespecificityoftheEgenetowardglucuronidase,preliminaryevidencesuggeststhattheeffectofthisgeneisalsotissue-specificsincenormalamountsofosmoticallyresistantglucuronidaseactivity(25-30%oftotal)arepresentinspleenhomogenates.Discussion.-ThepresentstudyshowsthataMendelianfactor,distinctfromtheglucuronidasestructuralgene,isrequiredfortheincorporationof8-glucuronidaseintothemembranesoftheergastoplasm.AnalteredformofthisfactorpresentintheYBRstrainpreventstheaccumulationofglucuronidaseinergastoplasmoflivercells.Theresultssuggestthattheintracellularcompartmentalizationofen-zymesmaybecontrolledbyfactorsotherthantheprimarystructureoftheenzyme.Toourknowledge,noothergeneticfactorwiththesameoranalogouspropertieshasbeenreportedinmulticellularorganisms.However,EdgarandWood'4havedemonstratedtheexistenceofgenesinphageT4Dwhoseproductsarerequiredforthemorphogenesisofthephageparticle.Ourresultshaveadditionalsignificancewhencomparedwiththeearlierreport'describingastructurallyalteredformofglucuronidasewhoserelativedistributioninthelysosomesandinthemicrosomeswaschanged.Theinterpretationfavoredinthatcasewasthatthealteredenzymeproteinpossessedadifferentrelativeaffinityforthebindingsitesofthelysosomesandtheergastoplasm.Itnowappearsthatinadditiontothestructureofglucuronidaseitself,aseparateformofnuclearcontrolexistsfortheintracellularlocalizationofliverglucuronidase.Themechanismofactionofthisadditionalcontrolcannotbedeterminedfromtheresultsofthepresentstudy.However,twoobviouspossibilitiespresentthem-selvesforconsideration:(1)theE+allelemaycodeforanenzymewhichactsuponnativeglucuronidasetoproduceanalteredformoftheenzymewhichisnowsusceptibletointegrationintheergastoplasmicmembrane;or(2)theE+alleledeterminesafactoressentialtotheconstructionofaproperglucuronidasebindingsiteintheergastoplasmicmembrane.Forthepresentwecannotchoosebetweenthesealternatives.However,itshouldbenotedthattheEgeneactsequallyuponbothstructuralformsofglucuronidase(Fig.4).Duringcellfractionation,themissingergastoplasmicenzymedidnotappearelsewhereinthecell.Furthermore,theresiduallysosomalactivityofYBRfell944PROC.N.A.S. GENETICS:GANSCHOWANDPAIGENwithintherangeoflysosomalactivitiesobservedamongnormalinbredstrainsofmice.'2Thesefactssuggestthateithertheliverhasafeedbacksystemwhichlimitstheproductionofglucuronidasetothenumberofavailableintracellularsites,orthatthecellshaveanefficientmethodofdisposingofexcesssolubleenzyme.Animportantfeatureofthebehavior

8 oftheEgeneisitsapparentspecificitytoward
oftheEgeneisitsapparentspecificitytowardglucuronidase.Nogrossupsetintheaccumulationofergastoplasmicproteinwasfound,asindicatedbythepresenceofwild-typeamountsofTCA-precipitablematerialandbythenormalelectropherogramofnicrosomalproteinfromYBRanimals.Furthermore,theclassicalmicrosomalenzymemarker,glucose-6-phosphatase,waspresentinwild-typeamountsandexhibitedatypicalmicrosomallocalizationaftercentrifugationofcellparticlesinasucrosedensitygradient.Evidently,ergastoplasmicglucuronidaseisunderanuclearcontrolmechanismdistinctfromotherproteinsofthisintracellularstructure.Ifthisspecificityshouldprovetoresideinspecialergastoplasmicbindingsites,theimpli-cationwouldbethatmost,ifnotall,ergastoplasmicproteinspossessspecialbindingsites.Wewouldliketoinferthatglucuronidase,andperhapsmanyotherproteins,areunderthecontrolofaspecialclassofgeneticelementswhichdeterminewhetheraproteincanexistinafunctionalstateataparticularintracellularlocation.Summary.-LivercellsofYBRmicelackergastoplasmicglucuronidase,butretaintheenzymeinlysosomes.ThelossofergastoplasmicenzymeisduetomutationofaMendeliangenewhichisdistinctfromtheglucuronidasestructuralgene.Thenewgene,symbolizedasE,doesnotappeartoaffectotherergastoplasmicproteins,andistissue-specific.Theresultsdemonstratethattheintracellularlocationofanenzymecanbedeterminedbyfactorsotherthantheprimarystructureoftheen-zyme,andthatseparategenescontrolthesefactors.Noteaddedinproof:WeapologizeforhavingunwittinglyomittedreferencetotherecentandelegantpaperofD.0.WoodwardandK.D.Munkres(thesePROCEEDINGS,55,872(1966)de-scribingnon-chromosomalgeneticvariantsofmitochondrialstructuralprotein.Thepresentrele-vanceoftheirworkisquiteobvious.*ThisinvestigationwassupportedbyUSPHSgrantsCA-04090andCA5016.tCurrentaddress:DepartmentofPharmacology,StanfordUniversitySchoolofMedicine,PaloAlto,California.1Paigen,K.,andW.K.Noell,Nature,190,148(1961).2Sidman,R.L.,andM.C.Green,J.Hered.,56,23(1965).3Paigen,K.,Exptl.CellRes.,25,286(1961).4Morrow,A.G.,E.M.Greenspan,andD.M.Carroll,J.Natl.CancerInst.,10,657(1949).5Paigen,K.,andR.Ganschow,inBrookhavenSymposiainBiology,no.18(1965),p.99.6deDuve,C.,B.C.Pressman,R.Gianetto,R.Wattiaux,andF.Appelmans,Biochem.J.,60,604(1955).7Bock,R.M.,andN.S.Ling,Anal.Chem.,26,1543(1954).8Davis,B.J.,Ann.N.Y.Acad.Sci.,121,404(1964).9Nimmo-Smith,R.H.,Biochim.Biophys.Acta,50,166(1961).'°Wattiaux,R.,andC.deDuve,Biochem.J.,63,606(1956).11Sumner,J.B.,Science,100,413(1944).12Ganschow,R.,andK.Paigen,Genetics,inpress.13Paigen,K.,thesePROCEEDINGS,47,1941(1961).14Edgar,R.S.,andW.B.Wood,thesePROCEEDINGS,55,498(1966).VOL.58,