HISTOPATHOLOGY manuscript) AND (non-edited HISTOLOGYskeletal muscle, muscle fiber types, single muscle fiber proteomics, mass spectrometry ... proteins or mitochondria, and the analysis

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

ONLINEFIRST

ThisisaprovisionalPDFonly.Copyeditedandfullyformattedversiónwillbemadeavailableatfinalpublication

Thisarticlehasbeenpeerreviewedandpublishedimmdediatelyuponacceptance.Articlesin“HistologyandHistopathology”arelistedinPubmed.

Pre-printauthor´sversion

ISSN:0213-3911e-ISSN:1699-5848

SubmityourarticletothisJournal(http://www.hh.um.es/Instructions.htm)

Fiber type diversity in skeletal muscle explored by mass spectrometry-based single fiber proteomics

Authors:Stefano Schiaffino, Carlo Reggiani and Marta Murgia DOI:10.14670/HH-18-170Articletype:REVIEWAccepted:2019-10-15Epubaheadofprint:2019-10-15

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

1

Fibertypediversityinskeletalmuscleexploredbymassspectrometry-basedsinglefiberproteomicsStefanoSchiaffino1,CarloReggiani2,3andMartaMurgia2,41VenetianInstituteofMolecularMedicine(VIMM),Padova,Italy2DepartmentofBiomedicalSciences,UniversityofPadova,Padova,3InstituteforKinesiologyResearch,ScienceandResearchCenterofKoper,6000Koper,Sloveniaand4Max-Planck-InstituteofBiochemistry,Martinsried82152,GermanyAddressforCorrespondence:StefanoSchiaffino,VenetianInstituteofMolecularMedicine(VIMM),ViaOrus2,35129Padova,Italy.E-mail:stefano.schiaffino@unipd.itShortrunningtitle:SinglemusclefiberproteomicsKeywords:skeletalmuscle,musclefibertypes,singlemusclefiberproteomics,massspectrometry

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

2

SummaryMammalianskeletalmusclesarecomposedofavarietyofmusclefiberswithspecialized

functionalproperties.Slowfibersaresuitedforlonglastingandlowintensitycontractileactivity,

whilevarioussubtypesoffastfibersareoptimizedtoproducehighforceandpowerevenwitha

significantfatigue.Thefunctionalspecializationofmusclefibersisbasedonselectivegene

expressionregulation,whichprovideseachfiberwithaspecificproteincomplement.

Therecentrefinementofsmall-scalesamplepreparation,combinedwiththedevelopmentof

massspectrometerscharacterizedbyhighsensitivity,sequencingspeedandmassaccuracy,has

allowedthecharacterizationoftheproteomeofsinglemusclefiberswithanunprecedented

resolution.Inthelastfewyears,thefirststudiesontheglobalproteomicsofindividualfibersof

differenttypeshavebeenpublished.Inthisshortreviewwediscussthemethodological

advancementswhichhaveopenedthewaytosinglefiberproteomicsandthediscoverypowerof

thisapproach.Weprovideexamplesofhowspecificfeaturesofsinglefiberscanbeoverlooked

whenwholemuscleormulti-fibersamplesareanalyzedandcanonlybedetectedwhenasingle

fiberproteomeisanalyzed.Thus,novelsubtype-specificmetabolicfeatures,mostprominently

mitochondrialspecializationoffibertypeshavebeenrevealedbysinglefiberproteomics.Inthe

sameway,specificadaptiveresponsesofsinglefiberstoagingorlossofneuralinputhavebeen

detectedwhensinglefiberswereindividuallyanalyzed.Weconcludethatthefibertype-resolved

proteomesrepresentapowerfultoolwhichcanbeappliedtoavarietyofphysiologicaland

pathologicalconditions.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

3

Introduction

Adistinctivefeatureofskeletalmuscleisthediversityandspecializationoffibertypes,which

allowmusclestofulfillavarietyoffunctionaltasksbyselectiverecruitmentofthemostsuitable

fibers:fatigueresistantslowfibersforlonglastingactivity,powerfulandfatiguablefastfibersfor

shortburstsofhighintensitywork.Biochemicalstudiesofselectedmuscleswithrelatively

homogeneousfibertypecomposition,whicharefoundinsomespecies,haveinitiallyprovided

somebasicinformationaboutthemolecularfeaturesofmuscletype-specificity.Forexample,the

seminalstudiesbyBarany(1967)basedonacomparativestudyoffastandslowmusclesfrom

differentspecies,revealedforthefirsttimetherelationshipbetweenmaximummyosinATPase

activityandthespeedofcontraction(seealsoMossandSolaro,2019).However,themixedfiber

typecompositionofmostskeletalmuscles,particularlyintwospecies,mouseandhuman,most

relevantforbiomedicalresearch,complicatestheinterpretationofbiochemicalstudiesatthe

wholetissuelevel.Dataderivedfromwholemusclestudiesreflectinfactboththeproportionof

thedifferentfibertypesandthevalueoftheparametersanalyzedineachfibertype.Inaddition,

wholemuscleanalysesalsoincludenon-musclecells,whichareaminorcomponentintermsof

globalproteincompositionbutamajorcomponentintermsofproteinspresentinthenucleus,as

myonucleirepresentlessthan50%oftotalnucleipresentinmuscletissue(Schmalbruchand

Hellhammer,1967).

Toaddressthisissueandidentifytheproteinsexpressedinspecificfibertypes,twoapproaches

weresubsequentlyused.Thefirstapproachisbasedonenzymehistochemistryand

immunohistochemistryanalysesofmusclesections.Thesecondapproachinvolvesthedissection

ofsinglefibersfromskeletalmusclesandtheanalysisoftheproteincompositionbySDS-PAGEand

Westernblotting.Acrucialstepinthestudyofmusclefibertypeswasthedemonstrationthat

skeletalmusclescontaindifferentmyosinheavychains(MYH)isoforms,whichcanbeidentifiedby

specificantibodiesandseparatedbyappropriateSDS-PAGEprotocols.MYHsthusbecamea

standardmarkertoidentifyfibertypesinmusclesectionsandinisolatedfibers.Thesestudiesled

tothecurrentviewthat,basedonMYHcomposition,mousemusclescontainfourmajorfiber

types,slowtype1andfast2A,2Xand2B,whilehumanmusclescontainthreemajorfibertypes,

slowtype1andfast2Aand2X(SchiaffinoandReggiani,2011).Additionalfibertypes,definedby

uniqueMYHisoforms,areconfinedtospecializedmuscles,suchascranialmuscles:forexample,

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

4

slow-tonicfibers,whichrespondtostimulationwithacontractureratherthanatwitch,are

presentinextraocularmusclesbutnotinmosttrunkandlimbmuscles(Rossietal.,2010).

Two-dimensionalgelelectrophoresiscombinedwithmassspectrometry(MS)allowedtoidentifya

significantnumberofproteins,uptoafewhundred,inskeletalmusclesamples(seeMurphyetal.,

2016;Capitanioetal.,2017).However,manymuscleproteins,e.g.MYHs,cannotbeanalyzed

becausetheydonotenterthegelinthefirstdimension(isoelectricfocusing,IF).Inaddition,two-

dimensionalgelelectrophoresis,whichrequiresabout100µgofprotein,cannotbeappliedtothe

studyofsinglemusclefibers,whichcontainaround1µgofprotein.Thislimitationalsoappliesto

otherapproachesinvolvingtheisolationofspecificmusclefiberstructures,suchascontractile

proteinsormitochondria,andtheanalysisoftheseproteinsbyMS.Sarcomericprotein-enriched

extractsofskeletalmuscles,derivedfrom5mgofmuscletissue,havebeenanalyzedbytop-down

targetedproteomicstodeterminealternativesplicingandpost-translationalmodificationsof

contractileproteins(Jinetal.,2019).Furthermore,gel-basedtechniquesarelaboriousandrequire

longhands-ontimepersample,thuspreventingmultiplexingoftheproceduretohighsample

numbers.

Theproteomicanalysisofisolatedsinglemusclefiberswasmadepossiblebyrecentadvancesin

proteomicstechniques,inparticularthedownscalingofthesamplesizethroughthedevelopment

ofshotgunMSapproaches(seebelow),whichallowtheidentificationofthousandsofproteinsin

asinglemusclefiber.Inthisshortreviewwediscussthemethodologicalbasisofthisapproachand

themainresultsobtainedinstudiesonmouseandhumanskeletalmuscles.

Methodologicalaspects

Amajorchallengeencounteredintryingtoapplymodernproteomicstechniquestosinglemuscle

fibersistheverylowamountofproteinscontainedinasinglemusclefiberor,inthecaseof

humanmuscles,inasegmentoffiberascanbedissectedfromabiopsysample.Theestimated

amountofproteininasinglemusclefiberisolatedfromtypicalmousemuscles,suchastheslow

soleusorthefastextensordigitorumlongus(EDL)isabout0.5μg,assumingacrosssectionalarea

of1000μ2,alengthof5mmandacylindricalshape.Astohumansamples,duetothickercross-

sectionalarea,similarcalculationsleadto0.5μgpermmlengthandthusapproximately1.5μg

withanaveragelengthof3mminabiopsytakenwithaBergstromneedle.Theseminutesample

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

5

amountsarenotsuitableforIF+SDS-PAGE-basedapproaches.Theproblemhasbeenaddressed

byoptimizinganin-solutiondigestionworkflow,wherebyallpreparationsteps,fromsamplelysis

totrypsindigestion,arecarriedoutinonebufferandconfinedtoasinglevessel.Thelatteris

subsequentlyalsousedforpeptidepurification,thusminimizingsampleloss(Kulaketal.,2014).In

ourstudiesweappliedthisapproachtosinglefibers,whichweremanuallyisolatedand

immediatelysnap-frozen(Murgiaetal.,2015and2017).Othergroupshaveusedcollagenase

digestionformusclefiberisolation(Langetal.,2018).

Asimplifiedworkflowofthesingle-fiberproteomicsusedinourstudiesispresentedinFig.1.An

indicationoftherelativeconcentrationofthevariousproteinsinthemusclefiberscanbederived

fromthecorrespondingIBAQ(Intensity-BasedAbsoluteQuantification)values,anoutputofthe

MaxQuantanalysissoftware(CoxandMann,2008;Tyanovaetal.,2016).IBAQvaluesaretheMS

signalintensitiesdividedbythenumberoftheoreticalpeptides,thusareroughlyproportionalto

themolarquantitiesoftheproteins.InourstudiesIBAQvaluesofindividualproteinswere

normalizedtotheIBAQvalueofα-skeletalactintoaccountforthevariablecross-sectionalarea

andlengthofthefibersanalyzed.

Afurthermethodologicalchallengeinsingle-fiberproteomicsisthewidedynamicrangeofprotein

expressioninmusclefibers,whichspansseveralordersofmagnitude,fromthehighlyexpressed

contractileproteins,suchasmyosinandactin,tolowabundanceproteins,suchastranscription

factors.InourMS-basedshotgunproteomicsworkflow,wemeasuretheabundanceofindividual

peptidesafterelutioninananoflowliquidchromatographygradientandelectrosprayionization

intothemassspectrometer.Thesamepeptidesarenotalwayspickedforsequencingbythemass

spectrometer,makingdetectionpartiallystochastic.Becauseofthedramaticdifferenceinthe

expressionlevelofsarcomericproteinscomparedtotherestofthemuscleproteome,peptides

fromafewhighlyabundantproteinsaremoreoftenselectedforsequencing,thusinterferingwith

theidentificationoflessabundantproteins.Asaconsequence,theoverallnumberof

identificationsandthecoverageofcomparablyunder-representedproteins(suchastranscription

factors)arereduced.Unlikemuscle,othertissuesandcelllineshavealargefractionofproteinsof

similarabundancerangeand,therefore,thequantificationofalargefractionoftheirproteomeis

technicallyeasier.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

6

AwaytoaddressthisissueandidentifylowabundanceproteinsistocombinetheMSanalysisof

singlefiberswiththatofa“library”,consistingofadeepproteomeofwholemusclehomogenates,

extensivelyfractionatedtoreduceproteomecomplexity.Suchalibrarycontainssequence

informationforpeptidesthataredetectedbutnotsequencedinthesinglemusclefibers.Using

the‘matchbetweenruns’featureoftheMaxQuantanalysissoftwareitispossibletotransfer

peptideidentificationsfromthedeepmuscleproteome,whereagivenpeptideismuchmore

likelytohavebeenfragmentedandsequenced,toasinglefiber,whereoftenonlytheintact

peptidehasbeenmeasured(Deshmukhetal2015).Asaresult,thenumberofquantifiedproteins

inmusclefibersincreases,allowingtheanalysisofalargerfractionofthesinglemusclefiber

proteome(Murgiaetal.,2015).Thenumberofpeptideswhichareidentifiedusing‘match

betweenruns’withthehelpofa“library”isaboutdoublethoseidentifiedwithoutmatching.

Anadditionalissuetobeconsideredisthatsomemuscleproteinfamilies,suchasMYHs,consistof

isoformswithextremelysimilaraminoacidsequences.Therefore,thenumberofpeptideswhich

canbeusedtodistinguishtheisoformsisonlyafractionofthetotalpeptidenumbermeasuredfor

theseproteins.Asaconsequence,theprecisequantificationofeachisoformcanbechallenging

usingthealgorithmsusedfortheanalysisofMSdatainthestandardmode,whichattributes

commonpeptidestothe“proteingroup”withthelargestnumberoftotalpeptidesidentified.To

avoidtheriskofincorrectattributionofMYHpeptidestothedifferentisoforms,onlyunique

peptides,i.e.peptideswhicharepresentonlyinagivenisoform,mustbeusedforquantification

ofMYHs.

ThecorrectidentificationandquantificationoftheMYHisoformshasprovidedspecificmarkersto

definetheproteinprofileofthedifferentfibertypes.Inagreementwithpreviousimmuno-

histochemicalandbiochemical(SDS-PAGE)studies,proteomicanalysesalsoshowthatasignificant

numberoffibershaveahybridMYHcomposition.However,wefoundthatbothhumanand

murinemusclescontainahighpercentageoffiberswithalargelypredominantisoform(morethan

80%ofthetotal)andthuscanbesafelyattributedtoagiventype.Onecanthuscomparethefiber

typeprofileofrelativelypuremusclefibertypes.Inthefollowingsectionsofthisreviewwe

highlightthepowerofsingle-fiberproteomicstorevealnovelaspectsofmouseandhumanmuscle

biology.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

7

Fiber-type-specificdifferencesinmitochondrialproteins

Thediscoverypowerofsingle-fiberproteomicsisclearlyillustratedbythestudyofmitochondrial

proteins.Alargenumberofmitochondrialproteinswasdetectedinthesingle-fiberproteome

datasetsofthefourfibertypespresentinmouseskeletalmuscles(Murgiaetal.,2015).The

comparisonofthesedatasetsshowsthatmostmitochondrialproteinsaremoreabundantintype

2Aand2Xfiberscomparedtotype1fibers,while,notsurprisingly,theglycolytic2Bfibersshowed

thelowestvalues.Forexample,therelativeexpressionlevelsof43proteinsfromcomplexI,III,IV,

andV(ATPase),whichwererepresentedinallourselectedfibers,showedthefollowing

coefficientsofOXPHOXquantity:1(type2A),0.91(type2X),0.63(type1),and0.21(type2B)

(Schiaffinoetal.,2015).Asimilardistributionofrelativevalueswasseenfortheproteinsofthe

TCAcycle,withatendencyfortype2Xfiberstoshowthehighestvalues.However,isocitrate

dehydrogenase(IDH)showedauniquepatternoffibertypedistribution,differentfromthatof

othercomponentsoftheTCAcycle.

IDHisanenzymethatcatalyzestheoxidativedecarboxylationofisocitrate,producingα-

ketoglutarateandCO2inatwo-stepprocess,whichinvolvesoxidationofisocitrateto

oxalosuccinate,followedbythedecarboxylation,formingα-ketoglutarate.Inmammals,IDHexists

intwoisoforms:IDH3isatetramercomposedbytwoα subunits,oneβ subunitandoneγsubunit

andcatalyzestheoxidativedecarboxylationofisocitratetoα-ketoglutaratecoupledwiththe

reductionofNAD+toNADH,whereasIDH2isahomodimerandcatalyzesthesamereactionusing

NADP+asacofactorinsteadofNAD+.Singlemusclefiberproteomicshasshownthatthe

distributionofIDH2andIDH3variessignificantlyaccordingtofibertypeinmurineskeletalmuscles

(Fig.2A).Inparticular,IDH2levelsdonotcorrelatewithOXPHOSlevels,astheyarehighestintype

1fibers,slightlylowerin2Aandmuchlowerin2Xand2Bfibers.Incontrast,IDH3α,aswellas

IDH3γ,ismuchlessabundantintype1comparedto2Xfibers,thushasadistributionsimilarto

thatofotherTCAcycleproteins.

ThedifferentialdistributionofIDHisoformswasconfirmedbyimmunohistochemistry(Fig.2B).

TheantibodiestoIDH2giveastrongsignalintype1and2Afibers,weakerin2Xandweakestin2B

fibers,whereasantibodiestoIDH3αstainstrongly2Xand2Aandveryweaklytype1fibers.The

observationshavebeenrecentlyconfirmedcomparingslowandfast2Afibersinmousesoleus

(Langetal.,2018).Thisdifferentialdistributionlikelyhasanimportantfunctionalcounterpart,as

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

8

IDH2supportsthereductionofNADP+toNADPHandthusprovidesareducedcofactorfor

glutathionesulphidereductaseandthioredoxinreductase,namelytheenzymesessentialfor

controllingthebufferingofsuperoxideviaglutathioneperoxidase,peroxiredoxinandsuperoxide

dismutase.TheabundanceofIDH2inslowfiberswouldthusimprovetheirabilitytocontrolredox

stateduringcontinuousmitochondrialATPgeneration(Fig.3).Thisinterpretationwasconfirmed

bythefindingthatnicotinamidenucleotidetranshydrogenase(NNT),theenzymethatcouplesthe

hydridetransferbetweenreducedNADHandNADP+toprotontranslocationacrosstheinner

mitochondrialmembrane,isalso10timesmoreabundantintype1fiberscomparedto2Xfibers,

inspiteofthegreaterabundanceofOXPHOSandmostothermitochondrialproteinsin2Xfibers

(Schiaffinoetal.,2015).IDH2andNNTarethemajormitochondrialenzymesinvolvedinNADPH

generation.Theirabundanceintheslowtype1fibersthussupportstheideathatthecontinuous

activityoftheslowfiberscausesagreaterROSproductionwhichiscounterbalancedbyagreater

abundanceofNADPHgeneratingenzymes.

Fiber-type-specificdifferencesintheadaptiveresponseofskeletalmuscles

ThediversityamongsinglemusclefibersisnotrestrictedtotheirspecializationinATP

regenerationortodifferentcontractileperformance.Thediversityinvolvesalsoregulatoryand

adaptiveprocesses.Thus,thesinglefiberanalysiscanrevealadaptivechangesthatmightremain

undetectedwhenmusclesampleswithheterogeneouscompositioninmusclefibersareanalysed.

Astrikingexamplecomesfromthestudyoffibertypespecificadaptationsduringagingofhuman

muscles(Murgiaetal.,2017).Singlefiberproteomicsconfirmedtheexpectedage-relateddecline

inmitochondrialenzymesforbothslowandfast2Afibers,butshowedunexpectedvariationsin

thecomplementofenzymesoftheglycolyticpathwayandofglycogensynthesisanddegradation.

Enzymesofglycolysisandglycogenmetabolismwerefoundupregulatedinslowmusclefibersof

elderlysubjects(average70years-old)comparedtoyoungsubjects(average24years-old),as

showninFig.4.Onecouldhypothesizethattheincreaseinglycolyticenzymesisacompensation

forthelossofmitochondrialfunction.Alternatively,itispossiblethatthemaineffectofthis

metabolicshiftisthecontrolofmusclefibertrophism.Inthisview,glycolyticintermediatescould

bedivertedtoprecursorsofnucleotides,aminoacids,andfattyacidstosustainmuscleprotein

synthesisinagedslowfibers.Thiscouldunderlietheobservedstrikingdifferenceinmusclefiber

trophismduringaging.Indeed,manypreviousstudiesshowthatthefastfibersofelderlysubjects

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

9

aredistinctlyatrophic,whereasslowfiberscharacteristicallymaintaintheirsizeduringaging.(e.g.

LexellandTaylor1991;Callahanetal2014).Interestingly,theabovementionedisoformsofIDH

showadifferentbehaviourwithaging,characterizedbyadecreaseinIDH2inslowfibersand

stablevaluesifnotincreasedinfast2Afibers.

Otherinterestingexamplesofhowsinglefiberproteomicscanrevealchangesinopposite

directionsofdifferentfibertypescoexistinginthesamemusclearegiveninthestudyofmurine

soleusmuscledenervation(Langetal.,2018).Singlefiberquantitativeanalysisrevealedopposing

regulationofSERCA2(codedbyATP2A2)inslowandinfast2Afibers.Removalofneuronal

innervationtendstoconvertslowfibersintofasterfibersandfastfibersintoslowerfibers.Inthe

murinesoleusmuscle,theapproximatecompositionrecognizesa50%-50%proportionofslowand

fast(mainly2A)fibers.Thus,theimpactofdenervationonaveryabundantproteinasthe

sarcoplasmic/endoplasmicreticulumcalciumATPasesmightremainundetectedwhenwhole

musclesareanalyzed,andonlythesinglefiberquantitativeanalysiscanrevealopposing

regulationofSERCA2(codedbyATP2A2)inslowandinfast2Afibers(Langetal.,2018).

Conclusionsandperspectives

Thedistinctivefeatureofsinglemusclefiberproteomicsisthat,incontrasttothetraditional

immunohistochemicalandelectrophoreticmethodsfocusedonspecificproteincomponents,this

techniqueprovidesaglobalandunbiasedportraitofthewholemyofiberproteinprofile.Forthe

firsttimeitisthuspossibletocompare,withinthesamefiber,proteinsassociatedwithdifferent

cellstructures,fromtheplasmamembranetomitochondria,sarcoplasmicreticulum,myofibrils

andnucleus,aswellastherelativedistributionofenzymesinvolvedinvariousmetabolic

pathways.Otheromicapproaches,suchastranscriptomics,havealsobeenappliedtosingle

musclefibers(seeChemelloetal.,2011,2019),howeverchangesingeneexpressiondonot

alwaysreflectchangesinproteinabundance(seeAndersenetal.,1997,1999),thusonlyprotein

levelscanbeusedtodrawmeaningfulphysiologicalcorrelations(seealsoSchiaffinoetal.,2019).

Singlemusclefiberproteomicscanbeadiscoverytooltodefinethefunctionofknownmuscle

proteins,asillustratedbythefiber-type-specificvariationsintheIDHisoformprofilereferredto

above,andtoidentifynewfiber-type-specificproteins.Indeed,dozensofproteins,previouslynot

consideredinskeletalmusclestudies,appeartobeselectivelyexpressedinoneoranotherfiber

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

10

type(seeTableS2inMurgiaetal.,2015,andTableS6inMurgiaetal.,2017).Proteomicdatathus

openanewwindowtoexploremusclefiberdiversity,bygeneratinghypothesesandsuggesting

experimentstovalidatethesehypotheses.Forexample,knockdownofIDH2,especiallywhen

combinedwiththelackofNNTthatoccursspontaneouslyincertainmousestrains(seeSchiaffino

etal.,2015),wouldbeexpectedtocauseincreasedROSproductioninslowtype1butnotinfast

type2fibersinresponsetoelectricalstimulationorforcedexercise(seeSchiaffinoetal,2019).

AlimitationofsinglemusclefiberproteomicsbasedonshotgunMSisthatitisunabletoresolve

alltheproteinvariantsderivedbyalternativesplicing.Top-downMS-basedproteomics,inwhich

intactproteinsareanalyzedratherthanpeptides,allowstobetterdistinguishproteinisoforms

derivedfromhomologousgenesandproteinsderivedbyalternativesplicingofthesamegene.

However,absolutequantificationofselectedproteinscanonlybeobtainedbytargeted

proteomics,addingpreciseamountsofspecificproteinslabeledbystableisotopestothesample

tobeexamined.Stableisotopeincorporationintroducesasmallmassdifferencetoidentical

peptidessothattheycanbedistinguishedbyMS.Differentmethodsforabsolutequantification

haveemergedoverthelastyearsincludingProteinEpitopeSignatureTags(PrESTs)(Zeileretal,

2012)andabsolutequantification(AQUA)(seeLindermannetal.,2017).Theyallrelyoneither

spikinginheavylabeledpeptidesorheavylabeledfulllengthproteins.

Differentproteomicapproachescanbealsobeusedtoanalyzethepost-translational

modificationsthatproteinsundergoinmusclefibers,asillustratedinarecentphosphoproteomics

studyontheeffectofexerciseinhumanskeletalmuscle(Hoffmanetal.,2015).Recent

methodologicaldevelopmentshaveshownthatsamplepreparationforphosphoproteomicscan

beconsiderablyscaleddownandmultiplexed,makingitamenabletotheanalysisofbiological

samplesofsmallsize(Humphreyetal.,2018).However,aphosphoproteomicsstudyofsingle

musclefibersisnotyetpossiblewithavailabletechniques,duetothetinyamountsofproteins

presentinsinglemusclefibers.Inconclusion,alikelyscenarioofthefuturedevelopmentof

musclefiberproteomicsisthatshotgunMSwillcontinuetoprovideaglobalpictureofthe

myofiberproteinprofileandwillbecomplementedbytop-downMSandtargetedproteomics

approachestodefineproteinvariantsgeneratedbyalternativesplicingandpost-translational

modifications.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

11

AfinalpointthatmustbekeptinmindisthatavailabledatabasesrequiredforMSsearchesdonot

includethelargeandcontinuouslyincreasingnumberofpreviouslyundetectedmicropeptides,

expressedfromlongnoncodingRNAs(lncRNAs)andcircularRNAs(circRNAs),manyofwhichhave

beenrecentlydiscoveredinskeletalandcardiacmuscle(MakarewichandOlson,2017;vanHeesch

etal.,2019).Thediscoveryofmicropeptidesfurtherincreasesthecomplexityofthe

“proteoforms”,i.e.themolecularformsofexpressedproteins,thatarepresentinanytissue,

includingskeletalmusclefibers(seeAebersoldetal.,2016).

Acknowledgements

WearegratefultoProf.MatthiasMannforhishelpandsupportinthegenerationofthe

proteomicdata.WealsothankDr.LeonardoNogaraforhishelpinthepreparationofthefigures.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

12

FigurelegendsFig.1.Graphicalillustrationoftheworkflowformassspectrometry-basedsinglemusclefiber

proteomics.Manualsinglefiberisolationandsnap-freezingisfollowedbyasinglebuffer/single

vesselprocedureforproteindigestion,followedbypeptidepurification.Liquidchromatography

connectedtonano-electrosprayprecedesMSanalysisofpeptides.Thecorrespondingspectrain

rawfilesareanalyzedusingthefreelyavailableMaxQuantcomputationalproteomicssoftware

platformandtherelatedPerseusframework.

Fig.2.Relativeabundanceofisocitratedehydrogenase2(IDH2)andIDH3αproteinsinthefour

fibertypespresentinmouseskeletalmuscle,asdeterminedbysingle-fiberproteomicsand

immunohistochemistry.A.Single-fiberproteomicsshowsthatIDH2hashighestlevelsintype1

andverylowlevelsintype2Xfibers;incontrast,IDH3αhashighestlevelsintype2Xandverylow

levelsintype1fibers.Type2AfibersshowrelativelyhighlevelsofbothIDH2andIDH3α,while2B

fibershavethelowestlevelsofbothisoformsduetotheirpoormitochondrialcontent. B.Fiber-

type-specificdistributionofIDH2andIDH3α intheslowsoleus(SOL)andfastplantaris(PLANT)

mouseskeletalmuscles,asdeterminedbyimmunohistochemistrywithspecificantibodies.Left

panels:Transversesectionswerestainedwithmonoclonalantibodiesspecificfortype1,2Aand

2Bmyosinheavychains(MYH)toidentifythedifferentfibertypes;type2Xfibersareunstained.

Middleandrightpanels:SerialsectionswerestainedforIDH2orIDH3αandco-stainedfor

dystrophin(green)tohighlighttheplasmamembrane.IDH2ismoreabundantintype1and2A

fibers,lessabundantin2Xandalmostundetectablein2Bfibers.Incontrast,IDH3αisexpressedat

muchhigherlevelsin2Xand2Afiberscomparedtotype1and2Bfibers.Thispatternof

expressioncloselycorrespondstotheMSvaluesderivedfromsingle-fiberproteomicsshowninA.

AdaptedfromSchiaffinoetal.,2015.

Fig.3.AlternativepathwaysoftheTCAcycleatthelevelofisocitratedehydrogenaserevealedby

single-fiberproteomics.Theyellowboxessuggestapossibleinterpretationtoaccountforthe

relativeroleofthetwopathways.TheNAD-dependentIDH3allowsthegenerationofmoreNADH,

thusmorefuelfortherespiratorychain(RC)andincreasedATPproduction,requiredbythemore

expensivemyosinmotorofthetype2Xfibers.TheNADP-dependentIDH2allowsthegenerationof

NDAPHrequiredbytheantioxidantmitochondrialsystemthatismoreabundantinthetype1

fibers.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

13

Fig.4.Age-dependentchangesinexpressionofglycolyticenzymesintype1and2Ahumanmuscle

fibers.Singlefibers,isolatedfromvastuslateralismusclebiopsiesofyounger(age22–27)and

older(age65–75)individuals,wereprocessedforsinglefiberproteomics.Thepercentagefold

change(older/youngerratio)isshownforeachproteininthetwofibertypes.Glycolyticenzymes

areindicatedbythecorrespondinggenenames.DatafromMurgiaetal.(2017).

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

14

ReferencesAebersoldR.,AgarJ.N.,AmsterI.J.,BakerM.S.,BertozziC.R.,BojaE.S.,CostelloC.E.,CravattB.F.,

FenselauC.,GarciaB.A.,GeY.,GunawardenaJ.,HendricksonR.C.,HergenrotherP.J.,HuberC.G.,IvanovA.R.,JensenO.N.,JewettM.C.,KelleherN.L.,KiesslingL.L.,KroganN.J.,LarsenM.R.,LooJ.A.,OgorzalekLooR.R.,LundbergE.,MacCossM.J.,MallickP.,MoothaV.K.,MrksichM.,MuirT.W.,PatrieS.M.,PesaventoJ.J.,PitteriS.J.,RodriguezH.,SaghatelianA.,SandovalW.,SchlüterH.,SechiS.,SlavoffS.A.,SmithL.M.,SnyderM.P.,ThomasP.M.,UhlénM.,VanEykJ.E.,VidalM.,WaltD.R.,WhiteF.M.,WilliamsE.R.,WohlschlagerT.,WysockiV.H.,YatesN.A.,YoungN.L.andZhangB.(2018).Howmanyhumanproteoformsarethere?Nat.Chem.Biol.14,206-214.

AndersenJ.L.andSchiaffinoS.(1997).MismatchbetweenmyosinheavychainmRNAandprotein

distributioninhumanskeletalmusclefibers.Am.J.Physiol.272,C1881-9.AndersenJ.L.,Gruschy-KnudsenT.,SandriC.,LarssonL.andSchiaffinoS.(1999).Bedrestincreases

theamountofmismatchedfibersinhumanskeletalmuscle.J.Appl.Physiol.(1985)86,455-460.

BárányM.(1967).ATPaseactivityofmyosincorrelatedwithspeedofmuscleshortening.J.Gen.

Physiol.50(6),Suppl:197-218.CallahanD.M.,BedrinN.G.,SubramanianM.,BerkingJ.,AdesP.A.,TothM.J.andMillerM.S.

(2014).Age-relatedstructuralalterationsinhumanskeletalmusclefibersandmitochondriaaresexspecific:relationshiptosingle-fiberfunction.J.Appl.Physiol.(1985).116,1582-1592.

CapitanioD.,MoriggiM.andGelfiC.(2017).Mappingthehumanskeletalmuscleproteome:

progressandpotential.Expert.Rev.Proteomics.14,825-839.ChemelloF.,GrespiF.,ZulianA.,CancellaraP.,Hebert-ChatelainE.,MartiniP.,BeanC.,AlessioE.,

BusonL.,BazzegaM.,ArmaniA.,SandriM.,FerrazzaR.,LavederP.,GuellaG.,ReggianiC.,RomualdiC.,BernardiP.,ScorranoL.,CagninS.andLanfranchiG.(2019).TranscriptomicanalysisofsingleisolatedmyofibersidentifiesmiR-27a3pandmiR-142-3pasregulatorsofmetabolisminskeletalmuscle.CellRep.26,3784-3797.

ChemelloF.,BeanC.,CancellaraP.,LavederP.,ReggianiC.andLanfranchiG.(2011).

Microgenomicanalysisinskeletalmuscle:expressionsignaturesofindividualfastandslowmyofibers.PLoSOne.6,e16807.

CoxJ.andMannM.(2008).MaxQuantenableshighpeptideidentificationrates,individualized

p.p.b.-rangemassaccuraciesandproteome-wideproteinquantification.Nat.Biotechnol.26,1367-1372.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

15

DeshmukhA.S.,MurgiaM.,NagarajN.,TreebakJ.T.,CoxJ.andMannM.(2015).Deepproteomics

ofmouseskeletalmuscleenablesquantitationofproteinisoforms,metabolicpathways,andtranscriptionfactors.Mol.CellProteomics.14,841-53.

HoffmanN.J.,ParkerB.L.,ChaudhuriR.,Fisher-WellmanK.H.,KleinertM.,HumphreyS.J.,YangP.,

HollidayM.,TrefelyS.,FazakerleyD.J.,StöckliJ.,BurchfieldJ.G.,JensenT.E.,JothiR.,KiensB.,WojtaszewskiJ.F.,RichterE.A.andJamesD.E.(2015).Globalphosphoproteomicanalysisofhumanskeletalmusclerevealsanetworkofexercise-regulatedkinasesandAMPKsubstrates.CellMetab.22,922-35.

HumphreyS.J.,KarayelO,JamesDEandMannM.(2018).High-throughputandhigh-sensitivity

phosphoproteomicswiththeEasyPhosplatform.Nat.Protoc.13,1897-1916.JinY.,DiffeeG.M.,ColmanR.J.,AndersonR.M.andGeY.(2019).Top-downmassspectrometryof

sarcomericproteinpost-translationalmodificationsfromnon-humanprimateskeletalmuscle.J.Am.Soc.MassSpectrom.Mar4.doi:10.1007/s13361-019-02139-0.

KulakN.A.,PichlerG.,ParonI.,NagarajN.andMannM.(2014).Minimal,encapsulatedproteomic-

sampleprocessingappliedtocopy-numberestimationineukaryoticcells.Nat.Methods11,319-324

LangF.,KhaghaniS.,TürkC.,WiedersteinJ.L.,HölperS.,PillerT.,NogaraL.,BlaauwB.,GüntherS.,

MüllerS.,BraunT.andKrügerM.(2018).Singlemusclefiberproteomicsrevealsdistinctproteinchangesinslowandfastfibersduringmuscleatrophy.J.ProteomeRes.17,3333-3347.

LexellJ.andTaylorC.C.(1991).Variabilityinmusclefibreareasinwholehumanquadriceps

muscle:effectsofincreasingage.J.Anat.174,239-249.LindemannC.,ThomanekN.,HundtF.,LerariT.,MeyerH.E.,WoltersD.andMarcusK.(2017).

Strategiesinrelativeandabsolutequantitativemassspectrometrybasedproteomics.Biol.Chem.398,687-699.

MakarewichC.A.andOlsonE.N.(2017).Miningformicropeptides.TrendsCellBiol.27,685-696.MossR.L.andSolaroR.J.(2019).Theenduringrelationshipbetweenmyosinenzymaticactivityand

thespeedofmusclecontraction.J.Gen.Physiol.151,623-627MurgiaM.,TonioloL.,NagarajN.,CiciliotS.,VindigniV.,SchiaffinoS.,ReggianiC.andMannM.

(2017).Singlemusclefiberproteomicsrevealsfiber-type-specificfeaturesofhumanmuscleaging.CellRep.19,2396-2409.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

16

MurgiaM.,NagarajN.,DeshmukhA.S.,ZeilerM.,CancellaraP.,MorettiI.,ReggianiC.,SchiaffinoS.

andMannM.(2015).Singlemusclefiberproteomicsrevealsunexpectedmitochondrialspecialization.EMBORep.16,387-395.

MurphyS.,DowlingP.andOhlendieckK.(2016).Comparativeskeletalmuscleproteomicsusing

two-dimensionalgelelectrophoresis.Proteomes.4,E27.RossiA.C.,MammucariC.,ArgentiniC.,ReggianiC.andSchiaffinoS.(2010).Twonovel/ancient

myosinsinmammalianskeletalmuscles:MYH14/7bandMYH15areexpressedinextraocularmusclesandmusclespindles.J.Physiol.588,353-364.

SchiaffinoS.andReggianiC.(2011).Fibertypesinmammalianskeletalmuscles.Physiol.Rev.91,

1447-531.SchiaffinoS.,ReggianiC.,KostrominovaT.Y.,MannM.andMurgiaM.(2015).Mitochondrial

specializationrevealedbysinglemusclefiberproteomics:focusontheKrebscycle.Scand.J.Med.Sci.Sports.25,Suppl4,41-48.

SchiaffinoS.,ReggianiC.andMurgiaM.(2019).Theroleofomicsapproachesinmuscleresearch.

In"OmicsApproachestoUnderstandingMuscleBiology".BurnistonJ.GandChenY.-W.(eds).Series"MethodsinPhysiology".SpringerUS.Inpress.

SchmalbruchH.andHellhammerU.(1977).Thenumberofnucleiinadultratmuscleswithspecial

referencetosatellitecells.Anat.Rec.189,169-175.TyanovaS.,TemuT.,SinitcynP.,CarlsonA.,HeinM.Y.,GeigerT.,MannM.andCoxJ.(2016).The

Perseuscomputationalplatformforcomprehensiveanalysisof(prote)omicsdata.Nat.Methods.13,731-740.

vanHeeschS.,WitteF.,Schneider-LunitzV.,SchulzJ.F.,AdamiE.,FaberA.B.,KirchnerM.,Maatz

H,BlachutS,SandmannC.L.,KandaM.,WorthC.L.,SchaferS.,CalvielloL.,MerriottR.,PatoneG.,HummelO.,WylerE.,ObermayerB.,MückeM.B.,LindbergE.L.,TrnkaF.,MemczakS.,SchillingM.,FelkinL.E.,BartonP.J.R.,QuaifeN.M.,VanezisK.,DieckeS.,MukaiM.,MahN.,OhS.J.,KurtzA.,SchrammC.,SchwingeD.,SebodeM.,HarakalovaM.,AsselbergsF.W.,VinkA.,deWegerR.A.,ViswanathanS.,WidjajaA.A.,Gärtner-RommelA.,MiltingH.,DosRemediosC.,KnosallaC.,MertinsP.,LandthalerM.,VingronM.,LinkeW.A.,SeidmanJ.G.,SeidmanC.E.,RajewskyN.,OhlerU.,CookS.A.andHubnerN.(2019).Thetranslationallandscapeofthehumanheart.Cell.178,242-260.e29.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

17

ZeilerM.,StraubeW.L.,LundbergE.,UhlenM.andMannM.(2012).AProteinEpitopeSignatureTag(PrEST)libraryallowsSILAC-basedabsolutequantificationandmultiplexeddeterminationofproteincopynumbersincelllines.Mol.CellProteomics.11,O111.009613.

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)

HISTOLO

GY AND H

ISTOPATHOLO

GY

(non-e

dited

man

uscri

pt)