RESEARCHARTICLEMutantCx30A88Vmiceexhibithydrocephalyandsexdependentb - PDF document

RESEARCHARTICLEMutantCx30-A88Vmiceexhibithydrocephalyandsex-dependentbehavioralabnormalities,implicatingafunctionalroleforCx30inthebrainNicoleM.Novielli-Kuntz*,EricR.Press1,2,*,KevinBarr,MarcoA.M.Prado1,2,3andDaleW.Laird1,2,ABSTRACTConnexin30(Cx30;alsoknownasGjb6whenreferringtothemousegene)isexpressedinependymalcellsofthebrainventricles,inleptomeningealcellsandinastrocytesrichinconnexin43(Cx43),leadingustoquestionwhetherpatientsharboringGJB6mutationsexhibitanybrainanomalies.Here,weusedmiceharboringthehumandisease-associatedA88VCx30mutationtoaddressthisgapinknowledge.BrainCx30levelswerelowerinmaleandfemaleCx30A88V/A88VmicecomparedwithCx30A88V/+andCx30+/+mice,whereasCx43levelswereloweronlyinfemaleCx30mutantmice.Characterizationofbrainmorphologyrevealedadisruptedependymalcelllayer,significanthydrocephalusandenlargedventriclesin3-to6-month-oldadultmaleandfemaleCx30A88V/A88VmicecomparedwithCx30A88V/+ HandlingEditor:StevenJ.ClapcoteReceived17June2020;Accepted3December2020 DepartmentofAnatomyandCellBiology,TheUniversityofWesternOntario,London,ON,CanadaN6A5C1.DepartmentofPhysiologyandPharmacology,TheUniversity

ofWesternOntario,London,ON,CanadaN6A5C1.RobartsResearchInstitute,TheUniversityofWesternOntario,London,ON,CanadaN6A5K8.*TheseauthorscontributedequallytothisworkAuthorforcorrespondence(dale.laird@schulich.uwo.ca) D.W.L.,0000-0002-4568-3285ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(https://creativecommons.org/licenses/by/4.0),whichpermitsunrestricteduse,distributionandreproductioninanymediumprovidedthattheoriginalworkisproperlyattributed. ©2021.PublishedbyTheCompanyofBiologistsLtdDiseaseModels&Mechanisms(2021)14,dmm046235.doi:10.1242/dmm.046235 DiseaseModels&Mechanisms valine(p.A88V)(Bosenetal.,2014;Kellyetal.,2019;Lukashkinaetal.,2017;Zhanetal.,2020).Theautosomaldominantp.A88VmutationinCx30isconsideredtobethecauseofHED2,becauseithasbeenidentifiedinnolessthansevenfamiliesencompassingatleast15individuals(Essenfelderetal.,2004;Smithetal.,2002;Zhangetal.,2003).Thep.A88Vmutantappearstoassembleintoleakyhemichannels,whichdisruptsCaandATPhomeostasisfoundintheskin(Essenfelderetal.,2004;Kuangetal.,2020).InordertounderstandtheconsequencesoftheA88Vmutationinvivo,Cx30-A88V

knock-inmiceweregenerated(Bosenetal.,2014).Thesemiceexhibitedoversizedsebaceousglandsandrelativelymildpalmoplantarhyperkeratosis(Bosenetal.,2014).Inarecentelegantstudy,Kuangetal.(2020)developedanovelanti-Cx30antibodythatcouldreversetheskinpathologymediatedbyleakyhemichannelsinhomozygousCx30-A88Vmutantmice(Kuangetal.,2020),raisingtheprofileofpotentialconnexin-targetedtherapeutics.HomozygousmiceharboringtheA88Vmutationexhibitedasecondpathologyoflow-frequencyhearingloss(Bosenetal.,2014).Intriguingly,thesesamemutantmicewereprotectedfromhigh-frequency,age-relatedhearingloss,suggestingthatdysregulatedCx30inthecochleacanprovideaphysiologicaladvantage(Bosenetal.,2014).FurthercharacterizationofthehearinglossprotectionfoundinCx30-A88Vmicerevealedthatage-dependentouterhaircelllosswasgreatlyreducedinmutantmice(Bosenetal.,2014;Kellyetal.,2019;Lukashkinaetal.,2017).Notably,Cx30isamplyexpressedinastrocytes,ependymalcells,leptomeningealcellsandbrainpericytes(Abudaraetal.,2014;Dereetal.,2003;Mazaréetal.,2018;Nagyetal.,2001)(seealsohttp://mousebrain.org/genesearch.html).ThisraisedthequestionastowhetherCx30-A88Vmutantmic

emightdevelopadditionalmorbiditiesduringagingrelatedtobraindevelopmentandfunction.Here,weinvestigatedCx30inthebrainsof3-to6-month-oldmaleandfemaleCx30-A88Vmutantmiceandcomparedthemwithlittermatecontrols.WefoundthathomozygousCx30-A88VmiceexhibitedlowerlevelsofCx30andsex-dependenteffectsonco-expressedconnexin43(Cx43).Homozygousmutantmicepresentedwithincreasedbrainweight,increasedventricularsizeandhydrocephalus.Strikingly,despitesimilarstructuralbrainchanges,homozygousfemalemutantmiceexhibitedworsebehavioraloutcomes,includingdeficitsinlearning.TheseexperimentsilluminatehowmutantCx30canimpactthemammalianbrainandcontributetopathology.RESULTSFemalemutantmiceexhibitgreaterreductionsinconnexinlevelscomparedwithmalesBothCx30andCx43areexpressedinthebrain,mostnotablyastheconnexinsforminggapjunctionsbetweenastrocytes,althoughCx43isfarmoreplentifulcomparedwithCx30(Nagyetal.,2001).ToinvestigatewhetherthepresenceoftheA88VmutationinCx30affectedeitherCx30and/orCx43levelsinthebrainsof3-to6-month-oldmutantmice,theirexpressionandlocalizationwereassessed.Inmales,Cx30mRNAlevelsweredecreasedincomparedwithwild-type(WT)mice(Fi

g.1A),whereasCx43mRNAlevelsweresimilarbetweengroups(Fig.1B).Infemalemice,however,Cx30mRNAlevelswerelowerinmiceandfurtherreducedinCx30comparedwithWTmice(Fig.1C).Furthermore,bothheterozygousandhomozygousmutantfemalemicedisplayedreducedCx43mRNAlevelscomparedwithWTmice(Fig.1D).Attheproteinlevel,Cx30waslowinCx30mice(Fig.2A,B),althoughCx43proteinexpressionwasunchangedbetweenthemalemousegenotypes(Fig.2A,C).Infemalemice,Cx30wasalsolessabundantinCx30mutantmicecomparedwithWTmice(Fig.2B,D),andCx43wasstatisticallylessabundantonlyinCx30A88V/+micecomparedwithWTmice(Fig.2D).HomozygousmutantmicedevelophydrocephalyTodeterminewhetherthealterationsinCx30andCx43affectedthebrainofagingmice,weexaminedbrainweightandgrossmorphologyinCx30+/+,Cx30A88V/+andCx30A88V/A88Vmaleandfemaleadultmice(Fig.3A).Analysisofcoronalbrainsectionsdemonstrated Fig.1.Cx30andCx43mRNAlevelsinwhole-brainlysatesfrommaleandfemaleWT,heterozygousandhomozygousmutantmice.(A-D)mRNAexpressionofCx30(A,C)andCx43(B,D)inwhole-brainlysateofmaleandfemalemice,respectively(******one-wayANOVA).ConnexinmRNAexpressionlevelsarenormalizedtoGAPDH. RESEARCHARTICLEDiseaseMod

els&Mechanisms(2021)14,dmm046235.doi:10.1242/dmm.046235 DiseaseModels&Mechanisms enlargedventricularareas(indicativeofvolumechanges)inbothmaleandfemaleCx30A88V/A88Vmicecomparedwithsex-matchedWTandheterozygotemutantmice(Fig.3B,C).Furthermore,bothmaleandfemaleCx30A88V/A88VmousebrainweightsweresignificantlygreaterthanthosefoundinCx30+/+orCx30A88V/+mice,despitesimilarbodyweight(Fig.3D,E).ThesefindingsclearlyindicatethatCx30A88V/A88Vmicehaddevelopedhydrocephaly.GiventhatCx30hasbeenshowntobeexpressedinependymalcellsliningthebrainventricles(Kunzelmannetal.,1999),andthesecellscontributetocerebralspinalfluidproduction,weassessedwhetherCx30andCx43levelsandlocalizationwerechangedinependymalcellsofmutantmice(Fig.4).Inbothmale(Fig.4A)andfemale(Fig.4B)mousebrains,Cx30wasdetectibleasgreenpunctaalongtheependymalcellsliningtheventricles(denotedbyasterisks),buttoamuchlesserextentthantheplentifullevelsofCx43.OwingtothelowlevelofCx30detectedandthelossofaclearlyintactependymalcellliningoftheventricles(Fig.4),itwasdifficulttodetermineconvincinglywhetherhomozygousmutantmicehadlessCx30comparedwithWTmice,althoughourwesternblot

analysismightsuggestthatthiswasthecaseinfemalemutantmice.However,itisimportanttorecognizethatthelevelofCx30mightalsobereducedinastrocytesandleptomeningealcellsofmutantmice,butthisisnotpossibletodiscernclearlyintheseimmunofluorescenceimages.ToassessthelocalizationofCx43andCx30inthewholebrainsofmaleandfemaleWTandmutantmice,coronalsectionsweredoubleimmunolabeledforconnexinandglialfibrillaryacidicprotein(GFAP)todemarcatethelocationofastrocytes(Fig.5).AlthoughCx43wasabundant,littleCx30immunoreactivitywasdetectedinastrocytesfrommale(Fig.5A)orfemale(Fig.5B)mice.Hence,theCx30localizationandexpressionpatternsinbraintissueweregenerallyconcurrentwiththereducedCx30mRNAandproteinlevelsseeninhomozygousmutantmice.BehavioralstudiesasameasureofcognitivefunctionGiventhatCx30miceexhibitedsex-dependentchangesinCx43andenlargedventriclessurroundedbyanapparentlydisruptedependymalcelllayer,itwasimportanttodeterminewhetherheterozygousmutantmice,whichmodelhumandisease,orhomozygousmutantmicepresentwithfunctionalbehavioraldeficits.Toassessgenerallocomotorbehavior,micewereexaminedusingtheopenfieldtest.Allmalemousegenotypestraveled

similardistancesover30mininanopenenvironment(Fig.6A,C).Interestingly,however,homozygousfemalemutantmicedisplayedmorevariabilityinlocomotoractivityandgreatertotaldistancetraveled,suggestingincreasedexploratorybehaviorinanewenvironment(Fig.6B,D). Fig.2.Cx30andCx43proteinlevelsinwholebrainlysatesfrommaleandfemaleWT,heterozygousandhomozygousmutantmice.RepresentativewesternblotsandproteinexpressionanalysisofCx30andCx43inwholebrainlysatesofmale(A,C)andfemale(B,D)mice(*one-wayANOVA).ImmunoblottingforGAPDHwasusedasaloadingcontrol.DataareexpressedrelativetoGAPDHexpression. RESEARCHARTICLEDiseaseModels&Mechanisms(2021)14,dmm046235.doi:10.1242/dmm.046235 DiseaseModels&Mechanisms TotestthebehaviorofCx30mutantmicefurther,theywerestudiedwiththeelevatedplusmaze(Fig.7).Allmalemicespentthemosttimeintheclosedarmandfarlesstimeineachoftheopenarms(Fig.7A).Notably,despitegreatervariabilitythanWTandmice,femaleCx30micespentsignificantlymoretimeintheopenarmsandlesstimeintheclosedarms(Fig.7B),supportingthenotionthatCx30femalemiceexhibitincreasedexploratorybehavior,eveninathreateningenvironment(openarms).Lastly,maleandfemalemice

weresubjectedtotheMorriswatermazetoassessspatiallearning.Forthecuedtest,wherethepositionoftheescapeplatformwasclearlymarkedwithamountedflag,bothmaleandfemalehomozygousmutantmicetooklongertoreachtheplatform(Fig.8A,B).However,inthecaseofthefemalehomozygousmutantmice,statisticalsignificancewasachievedonlywhencomparedwithheterozygousmutantmice,probablyowingtoawiderangeofvarianceintheWTcontrolgroup(Fig.8B).Duringthespatialacquisitionphase,onaverage,allmalegenotypesgraduallyreducedthetimetoescapeoverthe4daysoftesting.However,Cx30malesexhibitedlongerescapetimesatdays2and3comparedwithCx30andCx30(Fig.8C).Onday4,however,malehomozygousmutantmicedemonstratedescapetimessimilartothoseofWTandheterozygousmutants(Fig.8C).Infemalemice,bothWTandheterozygousmutantsexhibiteddecreasingescapetimeoverthe4daysoftesting(Fig.8D),whereashomozygousmutantsdidnotshowanyevidenceoflearning(Fig.8D).Duringthespatiallearningtask,allmicedemonstratedsimilarswimmingspeeds,suggestingthatdelayedlearningorlackoflearningwasnotattributabletopoorswimmingperformance(Fig.8E,F).DISCUSSIONAtpresent,nohomozygousGJB6missensemutationshavebeenidentifiedi

nthehumanpopulation.However,heterozygousmissensemutationsdoexistinpatientcohorts,withvariablepathologicalmanifestationsthatmostoftenaffecthearingandskinhealth,butitisnotatallclearwhetherandhowGJB6mutationsaffectthebrain,whereCx30iswidelyexpressed(Lamartineetal.,2000;TanandTay,2000).MutantmicelackingCx30demonstratereducedexploratoryactivity,withevidenceofincreasedanxiety- Fig.3.Maleandfemalehomozygousmutantmiceexhibitgreaterbrainweightandenlargedventricularvolumeindicativeofhydrocephaly.(A)MorphologicalrepresentationofWT,heterozygousandhomozygousmutantmalemice.(B,C)Histologicalrepresentationofcoronalbrainsectionsatcomparableposteriorpositionsinthethreegenotypesofbothmale(B)andfemale(C)mice.NotetheenlargedventricularareaofhomozygousmutantmicecomparedwithWTandheterozygousmice.SectionswerestainedwithHematoxylinandEosin.Scalebars:0.5mm,=3.(D,E)Male(D)andfemale(E)homozygousmutantmicedisplaysignificantlygreaterbrainweightthanWTandheterozygousmutantmicedespitesimilarbodyweights(****one-wayANOVA,Tukeysposttest). RESEARCHARTICLEDiseaseModels&Mechanisms(2021)14,dmm046235.doi:10.1242/dmm.046235 DiseaseModels&Mecha