Mosienko, V., Chang, A. J., Alenina, N., Teschemacher, A. G., & Kasparov, S. (2017). Rodents and humans are able to detect the odour of L-Lactate. PLoS ONE, 12(5), [e0178478]. https://doi.org/10.1371/journal.pone.0178478 Publisher's PDF, also known as Version of record License (if available): CC BY Link to published version (if available): 10.1371/journal.pone.0178478 Link to publication record in Explore Bristol Research PDF-document This is the final published version of the article (version of record). It first appeared online via Elsevier at https://doi.org/10.1371/journal.pone.0178478 . Please refer to any applicable terms of use of the publisher. University of Bristol - Explore Bristol Research General rights This document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/red/research-policy/pure/user-guides/ebr-terms/ RESEARCHARTICLE Rodents and humans are able to detect the odour of L-Lactate ValentinaMosienko1,AndyJ.Chang2¤,NataliaAlenina3,AnjaG.Teschemacher1, SergeyKasparov1,4* 1 SchoolofPhysiology,PharmacologyandNeuroscience,UniversityofBristol,Bristol,UnitedKingdom, 2 DepartmentofBiochemistryandHowardHughesMedicalInstitute,StanfordUniversitySchoolofMedicine, Stanford,UnitedStatesofAmerica,3 Max-Delbru¨ck-CenterforMolecularMedicine(MDC)Berlin-Buch, Berlin,Germany,4 BalticFederalUniversity,Kalinigrad,RussianFederation ¤ Currentaddress:DepartmentofPhysiologyandCardiovascularResearchInstitute,UniversityofCalifornia, SanFrancisco,UnitedStatesofAmerica *[email protected] a1111111111 a1111111111 a1111111111 Abstract a1111111111 a1111111111 L-Lactate(LL)isanessentialcellularmetabolitewhichcanbeusedtogenerateenergy.In addition,accumulatingevidencesuggeststhatLLisusedforinter-cellularsignalling.Some LL-sensitivereceptorshavebeenidentifiedbutwerecentlyproposedthattheremaybeyet anotherunknownG-proteincoupledreceptor(GPCR)sensitivetoLLinthebrain.Olfactory OPENACCESS receptors(ORs)representthelargestfamilyofGPCRsandsomeofthemareexpressed Citation:MosienkoV,ChangAJ,AleninaN, outsidetheolfactorysystem,includingbrain,makingtheminterestingcandidatesfornon- TeschemacherAG,KasparovS(2017)Rodents olfactoryLLsignalling.Oneofthe“ectopically”expressedORs,Olfr78inmice(Olr59inrats andhumansareabletodetecttheodourofL- Lactate.PLoSONE12(5):e0178478.https://doi. andOR51E2inhumans),reportedlycanbeactivatedbyLL.Thisimpliesthatbothrodents org/10.1371/journal.pone.0178478 andhumansshouldbeabletodetecttheLLodour.Surprisingly,thishasneverbeendemon- Editor:JohannesReisert,MonellChemicalSenses strated.HereweshowthatmicecandetecttheodourofLLinodourdetectionandhabitua- Center,UNITEDSTATES tion-dishabituationtasks,anddiscriminateitfrompeppermintandvanillaodours.Behaviour Received:January16,2017 oftheOlfr78nullmiceandwildtypemiceinodourdetectiontaskwasnotdifferent,indicating thatrodentsareequippedwithmorethanoneLL-sensitiveOR.Ratswerealsoabletouse Accepted:May12,2017 thesmellofLLasacueinanodour-rewardassociativelearningtask.Whenpresentedto Published:May25,2017 humans,morethan90%ofparticipantsdetectedasmellofLLinsolution.Interestingly,LL Copyright:©2017Mosienkoetal.Thisisanopen wasperceiveddifferentlythanacetateorpropionate—LLwaspreferentiallyreportedasa accessarticledistributedunderthetermsofthe pleasantsweetscentwhileacetateandpropionatewereperceivedasrepulsivesour/acid CreativeCommonsAttributionLicense,which permitsunrestricteduse,distribution,and smells.SubjectiveperceptionofLLsmellwasdifferentinmenandwomen.Takentogether, reproductioninanymedium,providedtheoriginal ourdatademonstratethatbothrodentsandhumansareabletodetecttheodourofLL. authorandsourcearecredited. Moreover,inmice,LLperceptionisnotpurelymediatedbyOlfr78.DiscoveryoffurtherLL- DataAvailabilityStatement:Allrelevantdataare sensitiveORmightshedthelightontheircontributiontoLLsignallinginthebody. withinthepaper. Funding:ThisworkwassupportedbyBBSRC:BB/ L019396/1,BB/K009192/1;andMRCMR/ L020661/1.Thefundershadnoroleinstudy design,datacollectionandanalysis,decisionto Introduction publish,orpreparationofthemanuscript. L-Lactate(LL)isproducedbyallcellsinthebody.Inthecentralnervoussystem(CNS)both Competinginterests:Theauthorshavedeclared thatnocompetinginterestsexist. neuronsandastrocytesproduceanduseLLbuttheabilitytostoreglycogenmakesastrocytesa PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 1/14 L-Lactateodourperceptioninrodentsandhumans particularlyimportantsourceofLLduringhighenergydemand[1,2],leadingtothehypothe- sisofastrocyte-to-neuronlactateshuttling.Accordingtothishypothesis,astrocytesproduce andreleaseLLwhichcanbetakenupandconvertedtoglucosebyneuronsduringperiodsof neuronalactivation[3,4].IthasrecentlybecomeevidentthatLLmayalsoactasasignalling moleculeviacognateG-proteincoupledreceptors(GPCR)[5].Oneofthese,knownasHCA1 (previouslyGPR81),hasnowbeenformallyrecognisedasareceptorforLL[6,7].HCA1is mainlyexpressedinadiposetissue,butlowlevelswerealsoreportedinvariouspartsofthe brain[8–10].SincemultiplestudiessuggestthatactivationofHCA1requiresveryhighLLcon- centrations[8,9],itisunlikelythatithasanyfunctioninthebrainunderphysiologicalcondi- tions.OurrecentresultspointedtotheexistenceofanotherLL-sensitiveG-coupledGPCRin s rodentbrain.ActingonthisputativeGPCR,LLevokesnoradrenalinereleaseandnoradrener- gicneurondepolarisationinthelocuscoeruleus[11].TheseeffectsofLLareconcentration- dependent,blockedbyD-Lactate,anddonotrequireLLentryintotheneuron.Identification ofthisLL-sensitiveGPCRisstillpending,raisinginteresttoanynovelGPCRwhichcanpoten- tiallymediateLLeffects.Inhumans,thereareabout800GPCRs,approximatelyhalfofwhich mediateolfaction(http://www.guidetopharmacology.org).Activationofolfactoryreceptors (ORs)leadstocAMPaccumulation,similarlytoG-coupledGPCRs.MostoftheORsare s exclusivelyexpressedinthesensoryneuronsofthemainolfactoryepitheliumbutectopic expressionofORshasalsobeenreported.Forexample,intestis,ORsarethoughttoplayarole inspermchemotaxis[12,13]buttheirrolesinothertissues,includingthebrain[14,15],are completelyunknown.OneoftheseectopicallyexpressedORsisthemouseolfactoryreceptor 78(Olfr78)whichhasratandhumanorthologues(Olr59andOR51E2,respectively[16,17]). Inthemouse,Olfr78wasdetectedinthebrainstemareapostremaandnucleustractussolitar- ius[16]andlaterfoundinsmoothmusclecellsinsmallbloodvesselsinavarietyoftissues[18, 19].Inaddition,itisalsoexpressedintypeIglomuscellsofthecarotidbody[20],prostate gland[21],colon,ovaries,testis,pancreas,andplacenta[22].HumanOR51E2andmouse Olfr78dose-dependentlyrespondedtoacetateandpropionate(Olfr78:EC =2.35mMfor 50 acetateand0.92mMforpropionate;OR51E2:EC =2.93mMforacetateand2.16mMfor 50 propionate[18]),andLL(Olfr78:EC =4mM[20]).Thewidedistributionofthisreceptor 50 impliesitsroleinvitalphysiologicalprocessesotherthanolfaction.Inthekidney,Olfr78was showntoregulateglomerularfiltrationrate,reninsecretion,andbloodpressure[18,19]while, inthecarotidbody,itwasproposedtoplayaroleinhypoxiadetection[20]. PresenceofOlfr78intheolfactorysystemanditssensitivitytoLLsuggeststhatrodentsand humansshouldbeabletodetectthesmellofLL.Surprisingly,thishasneverbeenconvincingly documented.Moreover,inthePubChemdatabase,informationaboutwhetherLLhasasmell iscontroversialwiththemostrecententrybeing“odourless”(https://pubchem.ncbi.nlm.nih. gov/compound/612#section=Top). Inthepresentstudy,weassessedwhetherrodentsandhumansare,infact,abletodetectan odourofLL,andtheimportanceofOlfr78inLLdetection.Weshowthatmice,ratsand humansareabletosmellLL.SinceOlfr78knockoutmicewereneverthelessabletodetectLL, theremustbemorethanoneLL-sensitiveORinthatspecies.IdentificationofalternativeLL signallingmechanismsmayshedlightontoLLtransmissionpathwaysoutsidetheolfactory system. Materialandmethods Animals AllproceduresinwildtypemicewereconductedinaccordancetotheguidelinesfromDirec- tive2010/63/EUoftheEuropeanParliamentontheprotectionofanimalsusedforscientific PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 2/14 L-Lactateodourperceptioninrodentsandhumans purposes,andtheexperimentsperformedinthisstudywereapprovedbytheSenateofBerlin EthicsCommittee(Landesamtfu¨rGesundheitundSozialesBerlin(LAGESO);studynumber G0300/13).TheexperimentsinratswereperformedaccordingtotheUKHomeOffice’sAni- mals(ScientificProcedures)Act,1986andapprovedbytheUniversityofBristolethicsreview committee(protocolnumber30/2875).ExperimentsperformedusingtheOlfr78null/reporter mouselinewereapprovedbytheInstitutionalAnimalCareandUseCommittee(IACUC)at theStanfordUniversitySchoolofMedicine(protocolnumber19968). Animalmaintenance 3and8maleC57Bl/6micewereusedforhabituation-dishabituationandodourdetection tasks,respectively.7homozygousOlfr78null[23]and9wildtypelittermatecontrolmiceona mixed129P2/OlaHsdandC57Bl/6Jbackgroundwereusedforanodourdetectiontask.All miceweremale,5–9monthsofage,group-housed,andmaintainedunderanormallight/dark cycle,withfreeaccesstofoodandwater. 12maleListerHoodedratsweighingbetween200and240g(Harlan)wereusedinan odour-rewardassociativelearningtask.Allratswerehandledandweigheddaily,andmain- tainedonareversed12-hourlight-darkcyclewithfreeaccesstofoodandwater.7daysafter thearrivalanimalswereputonarestricteddietfor10daysduringwhichtheyreceivedseven toninefoodpelletsperdaytomaintaintheirbodyweightat~85%oftheirfreelyfeeding weight.Suchadietregimewasmaintainedaswellduringtrainingandtestingdays.Thetask wasperformedduringthedarkphase. Odourdetectiontestinmice InordertotesttheabilityofwildtypemicetodetectthesmellofLL,wesubsequentlyintro- ducedpeppermint(40μloffreshlypreparedwater(distilled)extractfrompeppermintleaves), LL(40μloffreshlypreparedsolutionindistilledwater,1M,pH=7.4,L1750Sigma,(cid:21)98% purity),andurine(collectedatthedayoftestingfromfemalemicewithwhichthetestedmale micehadneverbeenincontact,10μl).TheodoursequenceintroducedtoOlfr78nullmiceand controlswaswater(40μl),LL(80μloffreshlypreparedsolutionindistilledwater,2M,pH~7.4, 71718Sigma,(cid:21)99%purity),urine(collectedatthedayoftestingfromfemalemicewithwhich thetestedmicehadneverbeenincontact,10μl),vanillaextract(40μlof1:10solutiondiluted indistilledwater),distilledwater(40μl).2MLLwaschosenforfurthertestinginOlfr78null miceandtheirlittermatessince2MLLinducedsimilarbehaviourresponsesincontrolmice onamixedbackgroundincomparisontoresponsesevokedby1MLLinC57Bl6/Nmice.30 minsbeforetesting,themalemousewasplacedintoafreshtestcage,andthenanodourwas introducedonapieceofWattmanpaper(2.5cmx2.5cm)inacornerofthecagefor3mins witha1minintervalbetweendifferentodours.Thetimemicespentsniffinganintroduced odorouspaperwasmeasuredbyanobserverblindtothetestedsubstances. Odourhabituation-dishabituationtestinmice ThisolfactiontestwasemployedtoassesstheabilitytodistinguishbetweenanodourofLL andothersmells,andanabilitytogetusedtoit.30minsbeforetesting,themousewasplaced intoafreshtestcage.Eachtestcagewasusedonlyonce.Duringthenext10minsadryand cleancottonstickwasintroducedintothecageinordertodecreaseunspecificsniffingnot connectedwiththeodour.LL(freshlypreparedsolutionindistilledwater,1M,pH=7.4, L1750Sigma,(cid:21)98%purity)wassubsequentlytestedwithpeppermint(freshlypreparedsolu- tionfrompeppermintleaves)andvanilla(dilutedindistilledwater1:100fromvanillaextract), anddistilledwaterasanegativecontrol.Eachsmellwasintroducedthreetimesfor2mins PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 3/14 L-Lactateodourperceptioninrodentsandhumans witha1mininter-trialintervalonacottonstickwhichwasdippedinthesolutionofthetested substance.Thetimemicespentsniffingdifferentodourswasmeasuredbyaresearcherblind totestedsubstances. Odour-rewardassociativelearninginrats Anodour-rewardassociativelearningtaskinratswasemployedtoidentifywhetherratsare abletousethesmellofLLtofindareward.Thistestwasperformedaccordingto[24]with modifications. Analuminiumoperantbox(80cmx80cmx50cm)wasused.Theinterioroftheboxwas spray-paintedmattblacktoavoidthepresenceofvisualcuesandtoprovideenoughcontrast forthemovement-trackingsoftware,Biobserve(Viewer2version2.2.0.91,BIOBSERVE GmbH,Bonn,Germany).Spongeswereallidentical,measuring8.5cmx7cmx4cm,witha 2cmholecutinthecentre,andwereonlyusedforonetestingdaywithanindividualanimal. Kellogg’sFrootLoopsusedinthestudywereshowntobeagoodreinforcementinoperant tasks[25].Allthesessionswererecordedbyavideocamerapositioneddirectlyabovetheoper- antboxtoallowobservationduringtrialsandsubsequentoff-lineanalysis. Fortheexperiments,onespongewasimpregnatedwitheithertestodourofLL(freshlypre- paredsolutionindistilledwater,1M,pH=7.4,L1750Sigma,(cid:21)98%purity)orcontrolodour ofalmond(1:100dilutedalmondextractindistilledwater)justbeforeeachtrial.Almondwas chosenforpositivecontrol,asitisknownthatratscansmellitandseemtohavenopreference noraversiontoit[26].Thespongewithconcealedrewardwasplacedinadifferentcornerof theoperantboxforeachtrialaccordingtoapredeterminedsequenceofnumbersfromaran- domnumbergenerator(1,2or3).Theratwasalwaysplacedinthesamecorner(0)ofthe box,nosefacingthewallatthestartoftrial,andtheremainingtwocornerswereoccupiedby emptyspongesimpregnatedwithdistilledwater.Intotal12ratswereusedforconditioning training,andeachratwastestedwithbothodours.Halfoftheanimalswerefirstconditioned withalmondodourand,followinga5-daybreak,withLLodour.Theotherhalfwerecondi- tionedinthereverseorder. Duringthefirsttrainingday,ratswereallowedtoexplorethearenafor5minbeforeasingle FrootLoopwasplacedontheflooroftheboxtofamiliarizetheratswiththereinforcement andtheexperimentalarena.Ratswerelefttoconsumethefoodandthenreturnedtotheir homecages,andtheprocedurewasrepeated(i.e.twotrainingepisodesforeachrat).During thesecondtrainingday,asimilarprotocolwasfollowedwiththeexceptionthatthree(odour- less)spongeswereplacedinthecornersofthebox,andduringthesecondtrialtheFrootLoop wasplacedintotheholeofaspongetoensurethattheratswereabletoretrievetherewardin thecourseoftheirnormalforagingbehaviour. OnthetestingdayafewFrootLoopswerecrushedtoapowderandsprinkledoverthefloor ofthewholeboxtopreventtheratsfrombeingguidedbythesmellofthehiddenFrootLoop duringthefollowingtrials.Eachsessionconsistedof6trials.Forthefirsttrialonly(introduc- torytrial0),thefoodrewardwasplacedonthetopofthespongesothattheanimalscould visuallylocateitandthistrialwasnotincludedintheanalysis.Thiswasnecessarybecausein thepreliminaryexperimentswefoundthatwhencrushedFrootLoopswerespreadaroundthe arena,ratsessentiallyneverfoundaconcealedrewardwithina5minuteperiod.Inthenexttri- alsthelatencyforacorrectresponse(timetakenforarattoretrievereward)anderrorswere recorded.Errorswerecountedasanose-pokeinawater-impregnatedsponge,orsniffingat thecorrectspongewiththeodourwithoutanose-pokeafterwards.Thesamespongeswere usedforrepeattrialswitheachratbutthepositionofthetargetspongewaschangedrandomly aftereverytrial.Afterfindingthefoodreward,ratswereallowedtoeatitbeforebeingremoved PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 4/14 L-Lactateodourperceptioninrodentsandhumans Table1. Keyoutcomesofthehumanodourdetectiontest. fractionofparticipantscharacterisedthesmellas* testsubstance ageofparticipants,years participantsdetectingthesmell pleasant unpleasant neutral L-Lactate 20.1±0.4 92.6%(87/94) 36.8%(32/87) 19.5%(17/87) 43.7%(38/87) acetate 20.5±0.6 95.3%(61/64) 6.6%(4/61) 49.2%(30/61) 44.3%(27/61) propionate 19.9±0.4 98.4%(60/61) 1.7%(1/60) 71.7%(43/60) 26.7%(16/60) pyruvate 20.0±0.4 96.8%(61/63) 52.5%(32/61) 24.6%(15/61) 23.0%(14/61) water 20.1±0.4 9.6%(9/94) 11.1%(1/9) 44.4%(4/9) 44.4%(4/9) *Percentageofparticipantswhoratedthepresentedodoursaspleasant,unpleasantorneutralwascalculatedfromthetotalnumberoftestsubjectswho detectedanodourinthejar. https://doi.org/10.1371/journal.pone.0178478.t001 fromtheboxandreturnedtotheirhomecage.Acut-offtimeof5minsand2–5minsofinter- trialintervalweremaintained. Odourdetectiontestinhumans TheprotocolforthisstudywasapprovedbytheUniversityofBristolEthicsCommittee (approval47121).Averbalconsenttobepartofthestudywasobtainedfromeveryparticipant. InordertotestwhetherhumanscandetectanodourofLL,werecruited94youngfemale andmalevolunteers(Table1),andaskedthemtoindicatethepresenceofanodourinfour jarscontainingvariouschemicalsorwater.Thetestwasperformedduringthreeindependent dayswithgroupsofmixedgender.Theorderoftheodourpresentationwasrandom,andthe participantswerenotinformedaboutwhatkindofodourstoexpect.100mlof1M(pH=7.4) freshlypreparedsolutionsindistilledwaterofLL(L7022Sigma,~98%purity),pyruvate (P2256Sigma,(cid:21)99%purity),propionate(P1880Sigma,(cid:21)99%purity)oracetate(S2889 Sigma,(cid:21)99%purity)wereputinwideodourlessglassjars,whichwerekeptclosedbetween tests.Distilledwaterservedascontrol.Theparticipantsweregivenaquestionnairetodeter- mine(1)whethertheycoulddetectanysmellinthejar(yes/no),(2)iftheyperceivedthissmell aspleasant,unpleasant,orneutral,and(3)whethertheyperceivedthesmellassweet,sour/ acid,bitter,neutral,or“hardtodefine”.Participantswereinstructedonhowtoperformthe testandthenexecuteditthemselves.Onlyoneparticipantwaspresentinthetestroomata time. Statisticalanalysis Resultsareexpressedasmean±SEM.Statisticalanalysiswasperformedbyone-ortwo-way ANOVA,orrepeatedmeasurementsANOVAwithBonferroni’scorrectionasapost-hoctest formultiplecomparisons,t-test,orFisher’sexacttestcalculatedfromacontingencytable (PRISM,GraphPad,SanDiego,CA).Toanalysereward-associatedlearninginratsallthe errordataweretransformedintocorrect/incorrectbinomialdataandevaluatedbycontin- gencyanalysistocomparedistributions(eitherchi-squaretestfortrendorchi-squaretestfor goodnessoffit).P<0.05wasconsideredsignificant. Results MiceareabletodetectanddiscriminatetheodourofL-Lactate First,wetestedinanodourdetectiontaskwhethermiceareabletosmellLL.Micespentsignif- icantlymoretimesniffingLL-impregnatedpapercomparedtowater(LL13.59±1.84secvs water4.84±0.64sec,p<0.01),whichwasclosetotheamountoftimetheyspentsniffing PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 5/14 L-Lactateodourperceptioninrodentsandhumans Fig1.MiceareabletodetectanddiscriminatetheodourofL-Lactate.A.Odourdetectiontask.Mice spentsignificantlymoretimesniffingL-Lactate(1M)thanwater,andthesameamountoftimeaspeppermint extract.Dataareshownasmean±SEM,n=8.*p<0.05,**p<0.01,***p<0.001vswater,pairedt-test.B. Odourhabituation-dishabituationtest.Onthefirstpresentation,micespentsignificantlymoretimesniffing L-Lactateincomparisontowaterandtothethirdintroductionofthepreviousodour(peppermint).Over severalintroductionsmicehabituatedtothesmellofL-Lactateandsniffeditforshortertimesduringthethird introductioncomparedtothefirstDataareshownasmean±SEM,n=3.#p<0.05vsthirdintroductionof previousodour,*p<0.05vsfirstwaterintroduction,pairedt-test.C.OdourdetectiontaskinOlfr78null mice.Olfr78nullmicespentthesameamountoftimeascontrollittermatessniffingL-Lactate(2M),orcontrol smells–vanillaandurineoftheoppositegender.Dataareshownasmean±SEM,n =7,n =9. Olfr78null WT https://doi.org/10.1371/journal.pone.0178478.g001 peppermint(LL13.59±1.84secvspeppermint10.68±1.79sec,p>0.05,Fig1A).Micespent11 timeslongersniffingapapercontainingurineincomparisontowater(urine52.12±3.85secvs water4.84±0.64sec,p<0.001,Fig1A),and4timeslongerincomparisontoLL(urine52.12 ±3.85secvsLL13.59±1.84sec,p<0.001,Fig1A).Thisisnotsurprisingbecauseforrodents urineoftheoppositegenderisoneofthemostpotentodorantsignals. Inanotherbehaviourparadigm,weintroducedeacholfactorystimulusthreetimes(water, peppermint,vanilla,andLL;Fig1B).Micehadlittleinterestinwaterbutwhenanyofthe odourswasintroducedforthefirsttime,theyspentasignificantlengthoftimesniffingthem (sniffingtimes:11.6±1.82secforpeppermint1,13.17±2.15secforLL1,and14.27±0.78secfor vanilla1vs3±1.02secforwater1,p<0.05).Whenanyoftheodorantswasintroducedrepeat- edly,sniffingtimesreducedfromthefirsttothethirdpresentation.Forexample,forLLdiffer- enceswereasfollows:secondvsfirstpresentation:1.63±1.17secvs13.17±2.15sec,p<0.01; thirdvsfirstpresentation:3.33±0.48secvs13.17±2.15sec,p<0.05.Micecoulddifferentiate theodourofLLfromthepreviouslyintroducedpeppermint(2.1±0.15secvs13.17±2.15sec, peppermint3vsLL1,p<0.05).ThesmellofLLwasalsodifferentiatedfromthesubsequently introducedvanillaasthesniffingtimeincreasedagain(lastpresentationofLL3.33±0.48secvs firstvanillapresentation14.27±0.78sec,p<0.01). Olfr78nullmicewerenotdifferenttowildtypecontrolsintheirabilitytodetecttheodour ofLL,andspentsimilaramountoftimesniffingodoursofLL,vanillaorurine(Fig1C).Partic- ularly,Olfr78nullmicespentsignificantlymoretimesniffingLL-,vanilla-orurine-impreg- natedpapercomparedtowater(water3.87±0.68secvsLL6.67±0.83sec,p<0.05;vanilla9.62 ±1.67sec,p<0.01;urine32.56±3.2sec,p<0.0001). RatscanusetheodourofL-Lactateasacue TotesttheabilityofratstodetecttheodourofLLweemployedanodour-rewardassociative learningtask(Fig2A).DuringthefirsttestinvolvingahiddenFrootLoop(trial1),ratstooka similaramountoftimetofindtherewardusingthesmellofLLoralmond(21.08±4.36secvs 26.98±7.78sec,LLvsalmond,p>0.05,Fig2B).Thefollowingtrials(2–5)revealedtheeffectof trialnumberonthelatencytofindtherewardbysmell(F(4,44)=4.610,P<0.01,Fig2B), whilethechoiceoftheodorantmadenosignificantdifference(F(1,11)=0.06604,P>0.05, Fig2B).Duringthelasttrial(5),ratsfoundtherewardmuchquickerthaninthebeginning PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 6/14 L-Lactateodourperceptioninrodentsandhumans Fig2.RatsareabletodiscriminatetheodourofL-Lactate.A.Schematicillustrationofanodour- rewardassociativelearningtaskinrats.Duringthetrainingdays,ratswerehabituatedtothearenaand rewardretrievalprocedure.Duringtheintroductorytrial(0)onthetestingday(notincludedinfurtheranalysis), aspongesoakedwiththetestodorantcarriedavisiblefoodreward.Duringconsequentfivetrialsonthe testingday,arandomlyplacedspongesoakedwiththetestsmelldispensedahiddenfoodreward. Outcomesofodourassociativelearningtaskinrats:B.Timetakenbyratstofindafoodrewardbythe smell(latency)wasnotdifferentbetweenL-Lactateandapositivecontrol,almond.Overthetrialsthelatency tofindafoodrewarddecreasedforbothodorants.Dataareshownasmean±SEM,n=12.*p<0.05, **p<0.01,two-wayANOVAwithBonferroni’scorrection.C.Distributionandmeanofcorrecttrialsforratsdid notdifferbetweenL-Lactateandalmond.Numberofobservedcorrecttrialswassignificantlyhigherthan expectedbychance.Dataareshownasmean±SEM,n=12.###p<0.001,##p<0.01vsmeanexpectedby chance,t-test. https://doi.org/10.1371/journal.pone.0178478.g002 (trial1vstrial5,p<0.5,Fig2B),irrespectiveoftheodorantcue.Next,weassessedhowmany ratssucceededtofindarewardwithoutanyincorrectnosepokes.LLwasequaltoalmondin thisregard(p>0.05,Fig2C),andtheprobabilityofacorrectchoicewashigherthanwhatmay beexpectedbychance(p<0.001foralmond,p<0.01forLL,Fig2C).Theprobabilityofran- domlychosencorrectresponseswascalculatedfromtheassumptionthatiftheratsdonotdis- tinguishanodourofeitherLLoralmond,theywouldhave1/3ofchancepertrialforacorrect nosespokebetweenthreesponges.Whencalculatedover5trials,thenumberofrandomly chosencorrectchoiceswouldbe1.67(Fig2C).Thedistributionofchancewasmodelled usingbinomialtheoremtocalculateexpectedproportionofratsforeachnumberofcorrect responses.Additionally,achi-squareforgoodnessoffittestedforadiscrepancybetween theobserveddistributionsforalmondandLLandtheexpecteddistributionbychanceand returnedverylowp-valuesforboth(p<0.0001forbothodourgroups).Altogether,thesedata indicatethattheodourconditioningenabledratstocorrectlyidentifylocationofthereward usingbothLLoralmondodoursasolfactorycues. HumansareabletodetecttheodourofL-Lactate YoungfemaleandmalevolunteerswereaskedtosmelljarscontainingsolutionsofLL,acetate orpropionatewhichwerereportedtoalsoactivateOlfr78invitro[18,20].Pyruvatewas includedasamoleculecarryingthesamecaloricvalueasLLandchemicallysimilartoLL,but whichwasnotpreviouslydescribedasanolfactorystimulant,noraligandforOlfr78.92.6%of theparticipantsindicatedthepresenceofasmellinajarcontainingasolutionofLL,36.8% and19.5%outofwhichcharacterisedthissmellaspleasantandunpleasant,respectively (Table1).Interestingly,96.8%ofthevolunteersdetectedanodorantinajarcontainingpyru- vate,againmainlycharacterisingitssmellaspleasant(52.5%vs24.6%,Table1).Tonote,the proportionofparticipantsindicatingpyruvateasapleasantodourwashighercomparedtoLL (pyruvate52.5%vs.LL36.8%,p<0.0001).Acetateandpropionatewerecharacterisedas PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 7/14 L-Lactateodourperceptioninrodentsandhumans Fig3.Gender-dependentdifferencesinL-Lactateodourperceptionbyhumans.A.Gender- dependentdifferencesinodourdiscrimination.Similarpercentagesoffemaleandmaleparticipantswere abletodiscriminateanodourinsolutionsofL-Lactate,pyruvate,acetateandpropionate.Numbersindicate amountofparticipantstestedwithasmell.B.Gender-dependentdifferencesin‘unpleasant’odour perception.L-Lactateandpyruvatewereperceivedasrepulsivebymorewomenthanmen(L-Lactate: 25.9%vs6.9%;pyruvate:33.3%and9.1%).Similarpercentagesoffemaleandmaleparticipantsratedthe odoursofacetateandpropionateasunpleasant.C.Gender-dependentdifferencesin‘pleasant’odour perception.Almosttwicemorementhanwomenratedthesmellofpyruvateasappealing(72.7%vs41%). SimilarpercentagesofmenandwomenratedtheodourofL-Lactate,acetateandpropionateaspleasant.D. Gender-dependentcharacterisationoftestodours.L-Lactate,pyruvate,acetateandpropionatewere ratedsimilarbymenandwomeninrespecttothedimensionoftheirsmell/taste.Mostparticipants, independentofgender,indicatedthemassweetandsour/acid,respectively.Lactateandwater:n(F)=64,n (M)=30;acetate:n(F)=47,n(M)=17;propionate:n(F)=41,n(M)=21;pyruvate:n(F)=40,n(M)=23. *p<0.05,Fisher’sexacttest. https://doi.org/10.1371/journal.pone.0178478.g003 odoroussubstancesby95.3%and98.4%,respectively,withalargemajoritycharacterising themasunpleasant(Table1). Therewasnosignificantdifferencebetweenmenandwomeninthefrequencyofdetection ofanyoftheodours(Fig3A).However,therewasadifferenceinhowLLodourwasperceived bymenandwomen,sincemanymorewomencharacteriseditasunpleasant(25.9%vs6.9%, p<0.05)(Fig3B).Curiously,menalsodidnotfindthesmellofpyruvateasunpleasantas womendid(33.3%vs9.1%,p<0.05).ThefrequencyofreportingLLasapleasantsmellwas notgenderdependentbutforpyruvatetheproportionofmenlikingitwassignificantlygreater (72.7%vs41%,p<0.05,Fig3C).Irrespectiveofgender,odoursofacetateandpropionatewere perceivedasanunpleasantbymostoftheparticipants(Fig3B). Perceptionofsmellishighlysubjective.Whenparticipantswereaskedtocharacterisetheir perceptionoftheodorants,LLandpyruvatewereoftendescribedasasweetsubstance,while acetateandpropionateweremostlycharacterisedassour/acid(Fig3D).However,morepartic- ipantsindicatedpyruvateodourassweetincomparisontoLL(LL27.6%vs52.5%pyruvate p<0.01).Onlyafewparticipants(9outof94)indicatedapresenceofodourinajarcontaining waterwhichapproximatestherateoffalsedetectioninthistest(p<0.0001forwatervsany othertestedodour,Fisher’sexacttest,Table1). Discussion LLisaremarkablemoleculewhichstandsoutfromthemyriadofchemicalswhichsurround usineverydaylife,becauseitispivotaltometabolismofallcellsandispresentinfoodsand PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 8/14 L-Lactateodourperceptioninrodentsandhumans bodysecretions[27–30].AfterthediscoveryofLL-sensitiveGPCRsinthebrain[6,7,11], LLwasrecognisedasapotentialsignallingmolecule,raisingtheinterestinanypossiblerecep- torsforthismolecule.Inhumans,ORsconstitutehalfoftheknownGPCRs(http://www. guidetopharmacology.org),andrepresentabiggroupofpotentiallyLL-sensitiveGPCRssome ofwhicharefoundoutsideoftheolfactoryepitheliumandprobablyimportantirrespectiveof theirroleinolfaction.Here,wetestedwhetherLLcanserveasanodorantinrodentsand humans,andassessedpotentialcontributionofOlfr78inLLodourperception.Tothisend,we performedseveralodourdetectionorodour-rewardassociativelearningtasksinrodentsand humans.InordertodelineatetheinvolvementofOlfr78insensingLL,weusedanOlfr78null mousemodelinanodourdetectiontask. HereforthefirsttimeweshowthatmiceareabletodetectthesmellofLLanddiscriminate itfrompeppermintandvanillaodours.Ratswereabletoperformanodour-rewardassociative learningtaskusingthesmellofLLasacue,whichworkedapproximatelyaswellasastandard odorant,almond. Whentestedinanodourdetectiontask,Olfr78nullmicehadthesamebehavioural responsesascontrolmicetothesmellofLL.Thisisnotentirelysurprisingbecausethereare onlyveryfewcaseswhenanablationof(oramutationin)oneoftheORswouldleadtoa changeinodour-evokedbehaviour.OneofsuchORsisodr-10inC.elegans,ablationofwhich leadstothelossofchemotaxisspecificallytodiacetyl,butnottoothersubstancesknownto causechemotaxisinthisspecies[31].Twofurtherexamplesofsuchreceptorsinmiceare trace-amineassociatedreceptor(TAAR)4and5,ablationofwhichresultedinlackofaversion topredatorurine[32]orlackofattractiontomurineurine[33],respectively. ThereisapossibilitythatthesensitivitytoLLwasnotcompletelylostbutonlyreducedin Olfr78nullmice.TestingresponsestolowerLLconcentrationsinodourdiscriminationtask mighthelptofurtherclarifythisissue. Ontheotherhand,odourdetectionandcodingistypicallycombinatorial.Thesameodour isusuallydetectedbymorethanonereceptorwhilethesamereceptormayberesponsiblefor thedetectionofdifferentodours[34,35].Basedonthesequencesimilarity,Olfr558(human orthologueOR51E1)isassignedtothesamesubfamilyasOlfr78anditisknownthatORsof thesamefamilyfrequentlydetectodorantswithsimilarstructures[35,36].Indeed,itwas shownthatOlfr558isactivatedbypropionate[37,38].ItisthereforepossiblethatOlfr558 mightbeabletointeractwithLLandcouldexplainwhyOlfr78deletioninmicedidnotpre- ventdetectionofLL.Interestingly,Olfr558isalsoexpressedoutsideoftheolfactorysystem [20]. TwootherORs,OR1G1andOR52D1,wererecentlyshowntobesensitivetopropionate [37,39],buttheyhavenotyetbeentestedfortheabilitytosenseLL.BLASTsearchusingthe cDNAsequenceofOlr59(ratanalogueofOlfr78,GenBankAY317512.1)asatagrevealed multipleolfactoryreceptorgeneswithmorethan60%similaritytoOlfr78.Onesuchageneis Olfr406,whichis96%identicaltoOlfr59attheDNAlevel.Thereisnoinformationonthe ligandsofthisreceptorbutitmightbeyetanotherORresponsibleforLLdetectioninOlfr78 nullmice. Whentestedinhumans,lessthan10%ofparticipantswereunabletodetecttheodourof LL.Interestingly,besidesLL,thechemicallyrelatedcompoundspropionateandacetatewere reportedtoactivatethehumanorthologueofOlfr78.However,subjectivelythesesubstances wereperceiveddifferentlybytheparticipants.Importantly,allchemicalswereneutralisedand theeffectsofprotonswhichcouldberesponsibleforsomeaspectsofsmellperceptionwas thereforeminimal.LLwasusuallyqualifiedasaratherpleasantodourwithasweetscent,and pyruvatewhichhasnotbeentestedasapotentialligandoftheORsdescribedabove,hadasim- ilarodourprofileasLL.ItislikelythathumanscandistinguishbetweenLLfrompyruvate PLOSONE|https://doi.org/10.1371/journal.pone.0178478 May25,2017 9/14