Psychoneuroendocrinology (2015) 56,110119Availableonlineatwww.sciencedirect.comScienceDirectj our nal homepage: www. el sevi er . com/ l ocat e/ psyneuenAdipocyteglucocorticoidreceptorsmediatefat-to-brainsignalingAnnetteD.deKloeta,b,c,,EricG.Kraused,MatiaB.Solomona,JonathanN.Flaka,b,KarenA.Scotta,b,Dong-HoonKime,BrentMyersa,YvonneM.Ulrich-Laia,StephenC.Woodsa,RandyJ.Seeleyf,JamesP.Hermana,aDepartmentofPsychiatryandBehavioralNeuroscience,UniversityofCincinnatiCollegeofMedicine,Cincinnati,OH45237,USAbGraduatePrograminNeuroscience,UniversityofCincinnati,Cincinnati,45237,USAcDepartmentofPhysiologyandFunctionalGenomics,UniversityofFloridaCollegeofMedicine,Gainesville,FL32611,USAdDepartmentofPharmacodynamics,UniversityofFloridaCollegeofPharmacy,Gainesville,FL32610,USAeDepartmentofPharmacology,KoreaUniversityCollegeofMedicine,Seoul,RepublicofKoreafDepartmentofSurgery,UniversityofMichiganSchoolofMedicine,AnnArbor,MI48109,USAReceived 14November2014;receivedinrevisedform23February2015;accepted4March2015KEYWORDSHypothalamicpituitaryadrenalaxis;Corticosterone;Adipose;Obesity;StressSummaryStress-related(e.g.,depression)andmetabolicpathologies(e.g.,obesity)areimportantandoftenco-morbidpublichealthconcerns.Hereweidentifyaconnectionbetweenperipheralglucocorticoidreceptor(GR)signalingoriginatinginfatwiththebraincontrolofbothstress andmetabolism.MicewithreducedadipocyteGRhypersecreteglucocorticoidsfollowingacute psychogenicstressandareresistanttodiet-inducedobesity.Thishypersecretiongivesrise todecitsinresponsivenesstoexogenousglucocorticoids,consistentwithreducednegativefeedbackviaadipocytes.Increasedstressreactivityoccursinthecontextofelevatedhypothal-amic expressionofhypothalamicpituitaryadrenal(HPA)axis-excitatoryneuropeptidesandinthe absenceofalteredadrenalsensitivity,consistentwithacentralciteofaction.Ourresultsidentify anovelmechanismwherebyactivationoftheadipocyteGRpromotesperipheralenergystorage whileinhibitingtheHPAaxis,andprovidefunctionalevidenceforafat-to-brainregu-latory feedbacknetworkthatservestoregulatenotjusthomeostaticenergybalancebutalsoresponses topsychogenicstimuli. 2015ElsevierLtd.Allrightsreserved.Corresponding author at: Physiology and Functional Genomics, University of Florida, College of Medicine, McKnight Brain Institute,100 S. Newell Drive (Bldg. 59, RM L4-162), Gainesville, FL 32611, USA. Tel.: +1 352 392 9236.Corresponding author at: Psychiatry and Behavioral Neuroscience, University of Cincinnati, 2170 East Galbraith Road ML0506, Cincinnati,OH 45237, USA. Tel.: +1 513 558 4813; fax: +1 513 558 9104.E-mail addresses: adekloet@u.edu (A.D. de Kloet), james.herman@uc.edu (J.P. Herman).http://dx.doi.org/10.1016/j.psyneuen.2015.03.0080306-4530/ 2015 Elsevier Ltd. All rights reserved.AdipocyteGRmediatefat-to-brainfeedback1111. IntroductionStressorsmobilizeenergyreservestoensuresurvivalunderenergeticallydemandingconditionsofrealorperceivedadversity(deKloetetal.,2005;Ulrich-LaiandHerman,2009).Aswouldthenbeexpected,thereisanintricaterela-tionshipbetweenthesystemsthatregulatemetabolismandthesystemsthatarestimulatedinresponsetostress.Acti-vationofthehypothalamicpituitaryadrenal(HPA)axisisaprimarycomponentofthemetabolicstressresponse,cul-minatinginthesecretionofglucocorticoids(corticosteroneinmice;cortisolinhumans)andconsequentredistributionoffuelsources(mobilizationofhepaticglucoseproduc-tion,enhancedadipocytedifferentiation).TheinterrelatedcontributionoftheHPAaxistostressandmetabolismisreectedinthelinkbetweenexcessglucocorticoidsandvis-ceraladiposity(e.g.,Cushingsdisease)(Masuzakietal.,2001;Pasqualietal.,2006),andbyevidenceforpatho-logicalHPAaxisactivityinpsychiatricpathologiessuchasdepression(Holsboer,2000)aswellasinmetabolicdisorderssuchasdiabetesandobesity(Masuzakietal.,2001;Pasqualietal.,2006;Rosmondetal.,1998).Furthermore,obe-sitypredisposesindividualstodevelopdepression(Robertsetal.,2003;Simonetal.,2006).StressactivationoftheHPAaxisiscontrolledbynega-tivefeedbackmechanisms,wherebyglucocorticoidsbindtocognatereceptorstoinhibitfurtherreleaseofACTH.Therearetwoknownreceptorsforglucocorticoids,themineralo-corticoidreceptor(MR)andglucocorticoidreceptor(GR).TheMRhashighafnityforglucocorticoidsandisexten-sivelyboundunderrestingconditions,evenatthenadirofthecircadianrhythm(DeKloetetal.,1998).Incon-trast,theGRisonlyextensivelyoccupiedathighcirculatingglucocorticoidlevels,andisthemajormediatorofnega-tivefeedback(Myersetal.,2012).Inrecentyearsithasbecomeapparentthatfeedbackcanbemediatedbymulti-plemechanisms.Forexample,fastfeedbackshut-offofCRHneuronsismediatedbynon-genomic,membraneglucocorti-coidsignaling,probablymediatedbytheGR(Evansonetal.,2010).Additionalregionsarealsoinvolvedinfeedbackinhi-bition,includingthehippocampus,medialprefrontalcortexandevennucleusofthesolitarytractneuronsinthehind-brain(Ghosaletal.,2014;McKlveenetal.,2013;Myersetal.,2012).Consequently,regulationofstressresponsesisadistributedprocessinvolvingmultiplebrainmechanisms.Althoughtheinter-relationshipbetweenstress-respondingandmetabolismisdocumented,theunderlyingmechanism(s)connectingthesystemsthatregulateenergystorageandthosethatregulatetheHPAaxisarenotclear.Thereisextensiveoverlapbetweenthebrainmecha-nismsregulatingstressresponsesandthosethatinuencemetabolismandthisislikelyfurthercomplicatedbyperipheralfactors.Inthisregard,ithasbeenhypothesizedthatafactorwithinadiposetissueplaysanimportantroleinmediatingtheinteractionsbycoordinatelyregulatingenergystorageandHPA-axisstressresponsiveness(Dallmanetal.,2003b;Laugeroetal.,2001).Consistentwiththishypothesis,theingestionofcomfortfoodsduringstressexposuresuppressesHPAaxisactivitybystimulatingrewardcircuitryinthebrain(Ulrich-Laietal.,2010),whiletheredistributionofadipositytowardincreasedvisceralstorescontributestotheattenuationofHPAresponding(Dallmanetal.,2003b;Laugeroetal.,2001;Pecoraroetal.,2004).WhileitisacceptedthatglucocorticoidsinhibittheirownsecretionviatheactivationofGRwithinspecicbrainregionsandinthepituitary,theexistenceofperipheralpopulationsofGRintissuessuchaswhiteadiposetissueraisesthepossibilityofreciprocalbody-to-brainfeedbacksignalsthatlinkmetabolicandneuralprocessinginthereg-ulationofkeystressresponses.ThepresentstudiesarebasedontherealizationthatGRishighlyexpressedinadipocytesandthereforeisinanidealpositiontomediatetheinteractionsbetweenstressandmetabolism.Toassessthispossibility,weinvestigatedtheroleofadipocyteglu-cocorticoidsignalinginenergymetabolismandHPAaxisactivityusingmicewithselectiveknockdownoftheGRinfatcells.Wedemonstratethatdirectactionofglucocorti-coidsonGRwithinadipocytesisanimportantmechanismforbothHPAaxisandmetabolicregulation.Thispathwaymayrepresentanimportantlinkbetweenobesityandpsy-chopathology.2.Materials andmethods2.1.AnimalsMicecontainingtheGRoxallele(Breweretal.,2003)werecrossedwithmicecontainingCrerecombinaseundercontroloftheadiponectinpromoter(Wangetal.,2010)togen-eratemice(C57BL/6129background)withreducedGRinadipocytes.Adultmaleandfemaleadipocyte-GRknock-downmice(KO)andlittermatecontrolsexpressingonlytheadiponectinCretransgene(andforanadditionalcon-trolexperimentcontainingonlytheGRoxallele[i.e.,noadiponectinCretransgene])werehousedonepercageona12h/12hlight/darkcycle.Micewere810wksoldattheinitiationofstudiesandwerefedeitherstandardrodentchow(HarlanTekladLM-485;3.1kcal/g;5%fat)orhigh-fatdiet(HFD;ResearchDietsD03082706;4.54kcal/g;40%fat).Unlessotherwisenoted,foodandwaterwereavailableadlibitum.AllprocedureswereapprovedbytheUniversityofCincinnatiInstitutionalAnimalCareandUseCommittee.2.2.HPAaxisTailbloodsampleswerecollectedduringthecircadiannadirandcircadianpeakofCORTsecretion.Foracutestress,micewereplacedinplasticrestrainersfor30min,andtailbloodsampleswerecollectedat0,30,60and120minafteronsetofrestraint.Forthedexamethasone-restraintchal-lenge,miceweregivendexamethasone(0.1mg/kgsc)orsalineand2h laterwereplacedinrestrainers,withbloodsampledasabove.Inordertodetermineadrenalresponsiv-itytoexogenousACTH,micewererstgivenahighdoseofdexamethasone(4ug/kgsc),whichwaspre-determinedtopreventendogenousCORTproductioninbothKOandCONmice.2hfollowingdexamethasoneadministration,miceweregivenalow0.01mg/kgdoseofACTHina0.5%BSAin0.1MPBSvehicle.Thisdosewaspre-determinedtocauseaCORTresponsethatisequivalentto50%ofthemaximalresponse.15minlatertailbloodsampleswerecollected112A.D.deKloetetal.intoEDTA-coatedtubesfortheassessmentofplasmaCORTlevels.2.3.AnalysisofplasmahormonesBloodsamplescollectedfromtail-bleedsaswellasfromter-minalexperimentswerekeptoniceuntilcentrifugationat6000rpmfor15minandplasmacollection.Plasmasampleswerestoredat80Cuntilassessmentofplasmahormonelevels.BloodglucosewasdeterminedusingaFreedomLiteglucosemeter.PlasmaCORTandACTHwereassessedbyRIA(Krauseetal.,2011).Plasmainsulin(duringadlibi-tumfeedingconditionsandsubsequenttoa16-hfast)andplasmaleptin(duringadlibitumfeedingconditions,subse-quenttoa16-hfastandduringa30minrestraintchallenge)wereassessedusingELISAkitsfromCrystalChem,Inc.Thedetectionlimitsofthesekitsare0.1ng/mland0.2ng/ml,forinsulinandleptinrespectively.PlasmaadiponectinwasassessedbythemousemetabolicphenotypingcenterattheUniversityofCincinnatiusinganELISAkitfromMillipore,Inc.(Cat.#EZRADP-62K).2.4.BodymassandfoodintakeMiceweregivenlow-fatchowuntil105daysofage.BetweenDay105andDay147,micewerefedaHFD.Duringthistime,bodyweightandfoodintakewereassessedevery34daysatthesametimeofday(24h afterlightson).2.5.BodycompositionandfatpadweightsBodycompositionwasdeterminedusingNMRtechnology(EchoNMR,Waco,TX)onunanesthetizedmiceaspreviouslydescribed(Taicheretal.,2003).Atthetimeofsacrice(Day147),distributionofadiposetissueintheepidydimal(eWAT),mesenteric(mWAT),retroperitoneal(rpWAT)andinguinal(iWAT)depotswasdeterminedbycarefullyremovingandweighingtheindividualfatpads.2.6.AdipocyteseparationTheadipocyteandstromalvascularfractionsofepidyd-malwhiteadiposetissue(eWAT;400mg)wereseparatedbycollagenasedigestion(200U/ml;60minat37Cunderconstantagitation),ltrationandcentrifugation.Floatingadipocytesandpelletedstromalvascularcellswerecol-lectedandRNAextracted.2.7.AdipocytemorphometryDeterminationofmeanadipocytesizeanddistributionofadipocytesizewasmodiedfrompreviousstudies(Kimetal.,2008).Imagesofadipocyteswerecapturedbyalightmicroscope(CarlZeiss,USA).Thecross-sectionalareaofadipocyteswasmeasuredonparafn-embeddedhematoxylin-andeosin-stainedsectionsofepididymalfatatamagnicationof200 byimageprocessingwithcus-tomizedsoftwarewritteninLabview9.0edition(NationalInstruments,TX,USA).Intotal,8001500adipocytesineachmousewereassessed.Thecolorimageswereconvertedtobinaryimagesbyadjustingthethreshold.Thecontourofadipocyteswasmadeclearwithpaintbrushfunctioniftheadipocyteswerenotclearlydiscriminatedbecauseobscureorbrokenoutlines.Thedistributionofadipocytesizewasdeterminedbyrelativefrequenciesofadipocyteshavingaspecicsizewithinasetinterval(250m2).Themeanadipocytesizeforeachmousewasdeterminedbyaveragingthecross-sectionalareaofallassessedadipocytesandthemeanadipocytesizesbetweenCONandKOmicewerecom-pared.Subsequently,theadjustedadipocytenumberofeachmousewasdeterminedbydividingweightofepididymalfatbythemeanadipocytesizeandtheadjustedadipocytesizewascomparedbetweenthegroups.2.8.RNAisolation,cDNAsynthesisandReal-timePCRGeneexpressionstudieswereconductedontissuesamplescollectedfrommaleCONandKOmicemaintainedonastandarddiet.RNAeasycolumns(Qiagen,Valencia,CA)wereusedtoisolateRNAfromtissuesamples(deKloetetal.,2011).DNAasetreatment(Qiagen,Valencia,CA)wasper-formedtominimizegenomicDNAcontaminationofRNAextracts.Forhypothalamicgeneexpressionanalysis,thehypothalamuswasdissectedfromthefrozenbrainsandsub-mergedin700lofRLTbufferfromtheQiagenRNAeasykitonthedayofRNAextraction.Forthehypothalamus,pitu-itaryandadrenals,RNAextractionandDNAasetreatmentprocedureswereperformedaccordingtothemanufacturersinstructions.AslightlymodiedprotocolwasusedforRNAextractionfromWAT,adipocytesandthestromalvascu-larfractionofadiposetissue.Asmallsample(