5-ht1a The neurobiology of depression and antidepressant action

5-ht1a,depression,antidepressant

NeuroscienceandBiobehavioralReviews37(2013)2331–2371

ContentslistsavailableatScienceDirect

NeuroscienceandBiobehavioralReviews

journalhomepage:http://wendang.chazidian.com/locate/neubiore

v

Review

Theneurobiologyofdepressionandantidepressantaction

PaulWillnera,?,JørgenScheel-Krügerb,CatherineBelzungc

a

DepartmentofPsychology,SwanseaUniversity,SingletonPark,SwanseaSA28PP,UKCenterofFunctionallyIntegrativeNeuroscience,UniversityofAarhus,Denmarkc

INSERM930andUniversityFrancois-Rabelais,Tours,France

b

article

info

abstract

Articlehistory:

Received24April2012Receivedinrevisedform26November2012

Accepted10December2012

Keywords:

UnipolarmajordepressionAntidepressantdrugsAffectiveneuroscienceHippocampusAmygdala

VentromedialprefrontalcortexAnteriorcingulatecortexNucleusaccumbensCaudatenucleusHabenulaStress

Treatmentresistance

Wepresentacomprehensiveoverviewoftheneurobiologyofunipolarmajordepressionandantide-pressantdrugaction,integratingdatafromaffectiveneuroscience,neuro-andpsychopharmacology,neuroendocrinology,neuroanatomy,andmolecularbiology.Wesuggestthattheproblemofdepressioncomprisesthreesub-problems:?rstepisodesinpeoplewithlowvulnerability(‘simple’depressions),whicharestronglystress-dependent;anincreaseinvulnerabilityandautonomyfromstressthatdevel-opsoverepisodesofdepression(kindling);andfactorsthatconfervulnerabilitytoa?rstepisode(adepressivediathesis).Wedescribekeyprocessesintheonsetofa‘simple’depressionandshowthatkindlinganddepressivediathesesreproducemanyoftheneurobiologicalfeaturesofdepression.Wealsoreviewtheneurobiologicalmechanismsofantidepressantdrugaction,andshowthatresistancetoantidepressanttreatmentisassociatedwithgeneticandotherfactorsthatarelargelysimilartothoseimplicatedinvulnerabilitytodepression.Wediscusstheimplicationsoftheseconclusionsfortheunder-standingandtreatmentofdepression,andmakesomestrategicrecommendationsforfutureresearch.

©2012ElsevierLtd.Allrightsreserved.

Contents1.2.3.

Introduction........................................................................................................................................Differentmechanismsfordepressionandantidepressantaction................................................................................Thepsychobiologyofdepression..................................................................................................................3.1.Thediathesis/stressmodel.................................................................................................................

3.1.1.Diathesis..........................................................................................................................3.1.2.Stress..............................................................................................................................3.1.3.Implications......................................................................................................................

3.2.EffectsofstressontheHPAaxisandhippocampus.......................................................................................

3.2.1.Neurotoxiceffectsofstress.......................................................................................................3.2.2.Morphologicalconsequences....................................................................................................

3.3.Frontalbraincircuitryunderlyingthesymptomsofdepression..........................................................................3.4.Whatdrivesthechanges?..................................................................................................................

3.4.1.Anhedonia........................................................................................................................

3.5.Mechanismsmediatingaffectiveinformation-processingbiases.........................................................................

3.5.1.Amygdalaandnucleusaccumbens...............................................................................................3.5.2.Thehabenula.....................................................................................................................

3.6.Failuretocope..............................................................................................................................

3.6.1.Stimulusappraisal................................................................................................................3.6.2.Learnedhelplessness.............................................................................................................

2332

233323342334233423352335233523362337233723382339233923392340234023402341

?Correspondingauthor.

E-mailaddress:p.willner@swansea.ac.uk(P.Willner).

0149-7634/$–seefrontmatter©2012ElsevierLtd.Allrightsreserved.http://wendang.chazidian.com/10.1016/j.neubiorev.2012.12.007

5-ht1a,depression,antidepressant

2332

P.Willneretal./NeuroscienceandBiobehavioralReviews37(2013)2331–2371

4.

5.

Rumination.................................................................................................................................3.7.1.Ruminationandrecovery........................................................................................................

3.8.Basalgangliainvolvementincoping.......................................................................................................3.9.Summary:stressanddepressioninconditionsoflowvulnerability......................................................................

3.10.Vulnerabilitytodepression:1.Kindling..........................................................................................3.11.Vulnerabilitytodepression:2.Predisposition...................................................................................

3.11.1.Animalmodelsofvulnerability...........................................................................................3.11.2.Thevulnerabilitybrainandthedepressedbrain.........................................................................

Mechanismsofantidepressantaction.............................................................................................................4.1.Potentiationofmonoaminetransmission.................................................................................................4.2.Post-transductionalmechanismsofantidepressantaction...............................................................................4.3.Neurogenesis...............................................................................................................................

4.3.1.Thepathwayfromsynapsetoneurogenesis.....................................................................................4.3.2.Functionalsigni?cance...........................................................................................................

4.4.HPAaxis....................................................................................................................................4.5.Whyareantidepressantsslowtoact?.....................................................................................................4.6.Treatmentimplications....................................................................................................................4.7.Treatmentresistance.......................................................................................................................

4.7.1.Pharmacokinetics,misdiagnosisandcomorbidity...............................................................................4.7.2.Parallelswithvulnerabilitytodepression.......................................................................................

4.8.Approachestothetreatmentofresistantdepression.....................................................................................

4.8.1.Newtargets.......................................................................................................................

Conclusionsandresearchimplications............................................................................................................References.........................................................................................................................................3.7.

234123422342234323432344234523452345234523462348234823492349235023502352235223532354235523562358

1.Introduction

Depressionisthecommonestpsychiatricdisorder.Itisthemostdisablingmedicalcondition,intermsofyearslosttodisability,anditisprojectedthatby2030depressionwillbetheforemostcon-tributortotheworldwideburdenofdisease(WHO,2008).Inthisreview,wefocusonunipolarmajordepressivedisorder,whichisde?nedinDSM-IV(AmericanPsychiatricAssociation,1994),asaconditioncharacterizedbythepresenceoflossofpleasureorinter-estinusuallypleasurableactivities(anhedonia),togetherwithanarrayofotherfeatures,includinganergia,changesinsleepandappetite,sadness,andsuicidalideation.Presentationsofunipo-larmajordepressivedisorder(whichweshallrefertoassimply‘depression’)canbeveryvariable,butthisfacthasnotfeaturedprominentlyintheliteraturethatweshallreview.

Depressionischaracterizedbyaprofoundlynegativeviewoftheworld,oneselfandthefuture(Beck,1967),andthisnega-tiveworld-viewhasbeenrelatedtonegativebiasesinattention,interpretationandmemory(MathewsandMacLeod,2005).Specif-ically,studiesofcognitiveprocessingindepressionhavereportedincreasedelaborationofnegativeinformation,dif?cultiesdisen-gagingfromnegativematerial,andde?citsincognitivecontrolwhenprocessingnegativeinformation,whichinteraliaexplainwhydepressedpeopleexperienceahighlevelofnegativeauto-maticthoughtsandpathologicalrumination(GotlibandJoormann,2010).Depressedpeopleareparticularlyvulnerabletonegativepsychologicalfeedback,whichhasadisproportionatelydisruptiveeffectonsubsequentperformance(Elliottetal.,1996).Inadditiontoanincreasedresponsetoaversiveevents,depressionisalsochar-acterizedbyadecreasedresponsetoanticipated(McFarlandandKlein,2008)oractual(Pizzagallietal.,2008;Chaseetal.,2010)rewards,andthisprovidesacognitiveexplanationofthecoresymptomofdepression,anhedonia.Thesetwocomplementarybiases,increasednegativityanddecreasedpositivity,arecentraltomuchoftherecentneurobiologicalliteratureondepression,becausetheyplaydirectlyintotwoofthemajorexperimentalmethodologies,functionalneuroimagingandanimalmodelsofdepression.

Sinceitsintroductionalmost50yearsago,themonoaminehypothesis(“some,ifnotall,depressionsareassociatedwith

anabsoluteorrelativede?ciencyofmonoaminesatfunctionallyimportantreceptorsitesinthebrain”,withthecorollarythatantidepressantsworkbycorrectingthesede?ciencies)haspro-videdthemajorneurobiologicalaccountofdepression.Indeed,untilrecently,itwastheonlysigni?canthypothesis,andwhileitspredominancehasbeentosomeextenteclipsedbynewercon-ceptsoverthepastdecade,itremainsthecasethatthemonoaminehypothesishasprovidedtheonlysigni?canttheoreticalframe-workforantidepressantdrugdevelopment,provingstubbornlyresistanttothenumerousandveryexpensiveattemptsbythephar-maceuticalindustrytobreakoutofthemonoaminestraitjacketwithdrugsthatactthroughnovelmechanisms.AssummarizedinFig.1,newerantidepressantsdifferfromtheolderdrugsindecreasingtheincidenceofunwantedsideeffectsand/ornarrow-ingtheneurochemicaltarget,ratherthanbyintroducingnovelmechanismsofaction.However,improvementsinbothantide-pressantresponseratesandtheslowonsetofclinicaleffect,requiringseveralweeksofchronictreatmenttoachievethefulleffect,havebeenminimal.Tolerabilityhasimproved,butdif-ferencesinef?cacyaresmallanddif?culttodemonstrate,andthereislittleevidencethatthenewerantidepressantsaremoreef?caciousthantheolderantidepressants.Indeed,oneoftheold-estantidepressants,thetricyclicclomipramine,remainsamongthemostef?cacious,alongsidetheserotonin-noradrenalinereup-takeinhibitor(SNRI)venlafaxine,theselectiveserotoninreuptakeinhibitors(SSRIs)sertralineandescitalopram,andtheatypicalantidepressantmirtazepine(Montgomeryetal.,2007;Ciprianietal.,2009).Antidepressantshaveconsistentlyshownonlymoder-ateresponserates,witharound30–40%ofpatientsbeingclassi?edasnon-responders,andthelatencyofclinicalonsetremainsstub-bornlylong(Trivedietal.,2006;HoltzheimerandMayberg,2011).Whileantidepressantef?cacyhasbeenclaimedforanumberofnon-monoaminergicdrugsthataremarketedforotherindications,andthefailureofsomenovelagentsmaytosomeextentinvolveincreasedregulatoryrequirements,therelativelackofprogressoverthepast50yearsisremarkable(Blier,2010;Baghaietal.,2011).

Inthispaperwepresentacomprehensiveoverviewoftheneu-robiologyofunipolarmajordepressionandantidepressantdrugaction,integratingdatafromaffectiveneuroscience,neuro-and

5-ht1a,depression,antidepressant

P.Willneretal./NeuroscienceandBiobehavioralReviews37(2013)2331–2371

2333

Fig.1.Antidepressantdrugdevelopment:newwineinoldbottles.The?guresummarizesthewaysinwhichthe?vemajornewerclassesofantidepressantdrugsrepresentre?nementsofthemechanismsintroducedintheoriginaldrugclasses,thetricyclicantidepressantsandmonoamineoxidaseinhibitors.

psychopharmacology,neuroendocrinology,neuroanatomy,andmolecularbiology,andfrompreclinicalandclinicalresearch.Insodoing,wedevelopaframeworkforunderstandingtheneurobiologyofdepression,whichalsoprovidesabasisforunderstandingthelimitedsuccessofresearchinantidepressantdrugdevelopment.InSection3,we?rstprovideadetailedaccountofa‘basic’psy-chobiologyofdepression,whichcentresontheeffectsofstressonneurobiologicalandpsychologicalfunctioninginindividualswhohavealowpredispositiontobecomedepressed.Wenextconsiderthemechanismsthatunderlievariousvulnerabilitiestodepression,andreviewevidencethatthesere?ectchangesinbrainfunctionthatresembleeffectsofstress,withtheresultthatdepressionismoreeasilyprecipitatedandlessstress-dependent.InSection4,wereviewrecentresearchonthemechanismsofantidepressantactionwhichdemonstratesthatantidepressantsessentiallycoun-teractandrepairtheeffectsofstress.Wealsoshowthatthefactorsimplicatedinresistancetoantidepressanttreatmentlargelyreca-pitulatethefactorsinvolvedinvulnerabilitytodepression,andarguethatantidepressantsareineffectiveundertheseconditionsbecausestressisofminorimportance.InSection5,wediscusstheimplicationsoftheseconclusionsfortheunderstandingandtreat-mentofdepression,andmakesomestrategicrecommendationsforfutureresearch.But?rst(Section2),weexplainwhyitisnecessarytogiveseparateconsiderationtotheanalysesofdepressionandantidepressantaction.

intheneuralbasesofdepressionandantidepressantaction,andtheactionofantidepressantscannotaccuratelybedescribedasreversingandnormalizingtheprocessesthataredysfunctionalinthedepressedbrain.

Theclearestevidencethatthebrainofanantidepressant-treatedpatientisnotinanormalstatecomesfromstudiesinwhichantidepressanteffectsareblockedbyacutedrugtreatments.AnexampleisshowninFig.2.Inthisstudy,afterdepressedpatientshadbeensuccessfullytreatedwithSSRIs,theywereadministeredalowdoseofthedopamine(DA)D2receptorblockersulpiride.Thiscausedaprofoundreturnofdepressedmoodinthetreatedpatients,

6.05.0

VAS rating

4.03.02.01.0

2.Differentmechanismsfordepressionandantidepressantaction

0.0

1

2

3 4 Time (h)

5 6 7

Afeatureofthemonoaminehypothesisofdepressionthathasgonelargelyunremarkedisthatitproposesasinglemechanismforbothdepressionandantidepressantdrugs:depressionresultsfromadecreasedfunctioninginNAand/or5HTwhichantidepressantsincreasebacktonormal.Thesamesymmetryisseeninmostofthemorerecenthypothesesthatwillbediscussedbelow.However,theassumptionofsymmetryisincorrect.Therearemanydifferences

Fig.2.Ademonstrationthattheantidepressant-treatedbrainisnotina‘normal’state.Volunteers(Con)ordepressedpatientswhohadrecoveredfollowingSSRItreatment(Dep)wereadministeredacutelyeitherplacebo(Plac)oralowdoseoftheDAD2receptorblockersulpiride(Sul).Sulpiridecausedareturnofseverelydepressedmoodinthepatientsbutnotincontrols.

AdaptedfromWillneretal.(2005).

5-ht1a,depression,antidepressant

2334

P.Willneretal./NeuroscienceandBiobehavioralReviews37(2013)2331–2371

butifanything,slightlyimprovedmoodinanon-depressednon-antidepressant-treatedcontrolgroup(Willneretal.,2005).Exactlythesameeffectisseeninstudiesofserotonin(5HT)ornoradren-aline(NA)depletion.Consumptionofamixtureofdietaryaminoacidsthatomitstryptophan,theprecursorof5HT,leadsrapidlytoa70–80%decreaseinplasmalevelsoftryptophan,andaconsequentdecreaseinbrain5HTlevels;andthetyrosinehydroxylaseinhibitoralpha-methyl-para-tyrosine(AMPT)causesaselectivedepletionofthecatecholaminesDAandNA.TryptophandepletioncausesaseverereinstatementofsymptomsinthemajorityofpatientsinremissionfollowingsuccessfultreatmentofdepressionwithSSRIs(Delgadoetal.,1990,1999).Similarly,AMPTprecipitatesasimilarrelapseindepressedpatientssuccessfullytreatedwithNAreuptakeinhibitors(NRIs)(Delgadoetal.,1993;Milleretal.,1996).However,neitherofthesemanipulationshasmajoreffectsonmoodinpeoplewhoarenotdepressedorathigh-riskfordepression(Ruhéetal.,2007).

Similar?ndingshavebeenreportedinanimalstudies.Both5HTdepletionandNAdepletionblocktheactionofSSRIsandNRIs,respectively,inanimalmodelsofdepression.However,asintheclinicalsituation,whilereversingtheeffectsofantidepres-sants,neithermanipulationproducesadepressivephenotypeinuntreatedanimals(LuckiandO’Leary,2004;Cryanetal.,2004;Yalcinetal.,2008).Chronicmildstress(CMS)isawell-validatedanimalmodelofdepression,basedonthelossofresponsivenesstorewardsbyanimalssubjectedtoavaryingscheduleofminorstressor(Willneretal.,1987;Willner,1997a).Inalargescaletranscriptomicstudyinmice,CMSinducedamolecularshiftincor-ticolimbicbrainareasthatwasonlypartlyreversedby?uoxetine:forexample,?uoxetinefailedtoblockoveraquarterofthechangesintheamygdalaandcingulatecortex,andalmostthreequartersofthechangesinthehippocampus,eventhoughthebehaviourofthemicehadreturnedtonormalaftertreatment(Surgetetal.,2009).

Itisevidentthatantidepressantsdonotnormalizebrainactivity:moodandbehaviourarerestoredtonormal,buttheantidepressant-treatedbrainisinadifferentstatefromthenon-depressedbrain.Thisisalsoevidencedbythehighrateofrelapseifantidepressantsaretoo-rapidlywithdrawnfollowingremissionofsymptoms(Kaymazetal.,2008).Thereforeitisnecessarytoexam-ineseparatelytheneurobiologyofdepressionandthemechanismsofantidepressantaction,ratherthancon?atingthesetwoissues.

3.Thepsychobiologyofdepression

3.1.Thediathesis/stressmodel

Individualswithinthepopulationvarygreatlyintheirvulnera-bilitytopsychiatricdisorders,includingdepression.Thisvariationisusuallyunderstoodwithinadiathesis/stressmodelthatconsidersseparatelyissuesofvulnerability(thediathesis)andprecipitation(thestress)(MonroeandSimons,1991).Twofeaturesofthismodelarecriticalforourpresentanalysis:asthediathesisincreases,thelevelofstressneededtoprecipitateanepisodeofdepressiondecreases,andtheoccurrenceofanepisodeofdepressionitselfincreasesthediathesisforfutureepisodes.

3.1.1.Diathesis

Weconsider?rstthenatureofthedepressivediathesis.Apre-disposition(ordiathesis)tobecomedepressedmayariseinavarietyofways,andatdifferentstagesofthelifecycle.Forexample,anumberofgeneticdiatheseshavebeenidenti?ed,withaheri-tabilityrangingfrom31to42%(Sullivanetal.,2000;Kendleretal.,2002,2006).Someearlylifeexperiencesarealsoknowntoincreasetheriskfordepression,particularlyinadequateemotionalcontactwithparents(RobertsonandBowlby,1952;Roy,1981;Slavichetal.,

2011)orchildhoodabuse(Kendleretal.,2002,2006;Widometal.,2007).Therefore,interactionsbetweenmultipleriskgenesandearlyenvironmentseemtoexplainalargepartofthevariability(KrishnanandNestler,2008;Caspietal.,2010),andmuchofthediathesisfordepressionislaiddowninearlychildhood.

Themechanismsbywhichtheseearlyexperiencesincreasetheriskofdepression(andotherpsychiatricdisorders)includenotonlybiologicalprocesses(discussedbelow)butalsopsychologicalandpsychosocialconstructsthatconverttransienttraumaticexperi-encesintolong-termvulnerabilities.Forexample,lossofaparentorapoorqualityofparentalcareleadstolowself-esteemandemo-tionalinstability(e.g.Akiskal,1984;AvagianouandZa?ropoulou,2008;Parsonsetal.,2010),andmaydecreasetheabilitytoformcloserelationships,andsodilutethequalityofsocialsupportavail-ableinlaterlife(e.g.Schoenfelderetal.,2011).Thereisevidencethatdistinctbutsubstantiallyoverlappingneuralnetworkssub-servedepressionandinsecureattachment(Galynkeretal.,2011).Evenhighlevelsofemotionsafterromanticloveandsexualactivi-tiesmayincreasetheriskforlaterdepressioninyoungadolescentgirls,asaresultofimmatureandinef?cientsocialcopingafterfail-uresatthisyoungage(Davilaetal.,2009).Earlylifeexperiencesalsodeterminecharacteristicstylesofprocessinginformationinrelationtotheself.Forexample,thenegativethinkingthatcharac-terizesthedepressedpersonisthoughttore?ecttheactivationofanegative‘cognitiveschema’,learnedthroughadversechildhoodexperiencessuchasrejection,criticism,orlivingwithadepressedparent(Beck,1967).Althoughdepressivecognitionsaretoalargeextentstate-dependentandtheirroleasvulnerabilitieshasbeenquestioned,thereisnowgoodevidencefortheexistenceofcog-nitivediathesesfordepression(Alloyetal.,1999;MathewsandMacLeod,2005;Roiseretal.,2012),whichmayonlybeapparentwhenadepressiveschemaisactivated:forexample,girlsatriskfordepressionbecausetheirmotherswererecurrentlydepressedattendedselectivelytonegativefacialexpressionswhensubjectedtoadepressivemoodinductionprocedure,butnototherwise(Joormannetal.,2007).

Personalityfactors,whicharemoderatelyheritable,butalsore?ectearlyexperience,interactwithbothcognitiveandsocialfactorsintheetiologyofdepression(Compassetal.,2004).Muchofthein?uenceofbothgeneticsandearlytraumaticeventsonchronicdepressivesymptomatologyismediatedthroughtheper-sonalityfactorofneuroticism(KendlerandGardner,2011),whichisoneofthestrongestriskfactorsfordepression(EnnsandCox,1997;ChristensenandKessing,2006).Early-onsetdepressioninparticular(?rstepisodebeforetheageof30),ischaracterizedbyahigherlevelofneuroticismandahigherprevalenceofcomorbidpersonalitydisorders,butlowerexposuretostressfullifeeventspriortoonset,relativetoindividualsexperiencingalater?rstepisode(Bukhetal.,2011).Likedepression,ahighlevelofneuroti-cismisalsoassociatedwithanegativeinformation-processingbias(Chanetal.,2007).Highlevelsofnegativeemotionality(aconstructcloselyrelatedtoneuroticism)hasbeenshowntolead,inyoungpeople,totheformationofdysfunctionalattitudesandothercog-nitivevulnerabilities(LakdawallaandHankin,2008;Joineretal.,2005),whilelowlevelsofpositiveemotionality(pleasurecapac-ity)increasevulnerabilitytodepressionbyreducingsocialsupport(WetterandHankin,2009).Neuroticindividualsathighgeneticriskformajordepressionalsohaveanelevatedtendencytointer-actwithothersinwaysthatgeneratestress,andtheybearsomeresponsibilityforthestressfullifeeventsthattheyencounter(Hammen,2006;Kendleretal.,2006).

Personalityfactorsandinparticular,thepersonalitystyleslabelled‘sociotropy’and‘autonomy’,mayalsoaccountformuchofthevariabilityinthesymptomatologyofdepression,includ-ingtheclinicalpresentationsofautonomous(or‘endogenous’)andreactivedepression.Autonomouspeopleobtainpleasurefrom

5-ht1a,depression,antidepressant

P.Willneretal./NeuroscienceandBiobehavioralReviews37(2013)2331–2371

2335

‘doing’andreachinggoals.Theyhaveautonomous(‘endogenous’)depressionscharacterizedbyinterpersonaldistanceandhostility,hopelessness/suicidality,feelingsoffailure,andanhedonia:havingfailedintheirowntermstheylosetheirmotivation,andblamingthemselvesfortheirfailure,theyminimizetheimportanceofenvi-ronmentalprecipitants.Sociallydependentpeoplerelyfortheirsatisfactionontheapprovalofothers:theirdepressionsarechar-acterizedbyinterpersonalsensitivity,guiltandself-blame,anxietyandrumination.Theirdepressionsare‘reactive’:theywillshowtemporarymoodimprovementsiftheyreceiveastronginputofattentionandreassurance;andbecauserelationshipswiththeoutsideworldarecentraltothem,theeventsthatprecipitatedepressionaremuchmoreobvious,andmaybeexaggeratedintheattempttoregainloveorattention(Beck,1983;Willner,1985;Robinsetal.,1997).Contrarytoawidespreadbelief,‘endogenous’and‘reactive’depressionsaredistinguishedbytheirpresentationratherthanbythepresenceorabsenceofprecipitants:indeed,endogenousdepressionismorestress-sensitive,notless(HarknessandMonroe,2006).

3.1.2.Stress

A?rstepisodeofdepressionmanifestsagainstthebackgroundofalevelofvulnerabilitythatisdeterminedbythesegeneticandexperientialfactors.Theprecipitant(orstressor)foradepressiveepisodecouldbeaninternalevent,suchasahormonalchallengeoratraumaticheadinjury.Morecommonly,theonsetofdepres-sionisprecipitatedbyexternalevents.Thelikelihoodofenteringadepressiveepisodeisgreatlyincreasedfollowingamajoradverselifeeventsuchasbereavement(recentevidencesuggeststhatinterpersonallosseventsareparticularlypowerfulprecipitantsofdepression:FarmerandMcGuf?n,2003;Slavichetal.,2010,2011),orinrelationtoanaccumulationofchronicminorstressessuchasunemployment,poverty,familydisharmonyorlivingwithsev-eralpre-schoolchildren(BrownandHarris,1978;HarknessandMonroe,2006).Thesetwotypesofprecipitatingeventsrespectivelyformthebasisoftwoofthemajoranimalmodelsofdepression,learnedhelplessness(Seligman,1975)andCMS(Willner,1997a).Bereavementhastraditionallyprecludedadiagnosisofdepres-sion,butthisexclusionhasbeenquestioned,onthegroundsthatbereavementrelateddepressionisinallrespectssimilartootherdepressiveepisodes(Corrubleetal.,2011),anditislikelytobereversedinDSM-V(Zisooketal.,2012).

Whilethecontributionofearlylifestresstotheetiologyofthedepressioniswellestablished,theroleofstressduringadulthoodhasbeendebated,becauseitappearsthatmanyinstancesofdepres-sionsarenotprecipitatedbystress.However,thereisasimpleresolutionofthiscontroversy.Itfollowsfromthediathesis/stressconceptthatapersonwhohasaweakdepressivediathesiswouldonlysuccumbtoanintensestress,whereasapersonwithastrongdepressivediathesismaysuccumbtominorortrivialstresses.Forexample,inalargetwinstudy,theoddsrationforanassociationbetweenseverelystressfullifeeventsandtheonsetofa?rstepisodeofdepressionwasfourtimesgreaterinwomenwiththelowestlevelofgeneticrisk(dizygotictwinswhoseco-twinhadnohistoryofdepression)thaninwomenwiththehighestlevelofgeneticrisk(monozygotictwinswhoseco-twindidhaveahistoryofdepres-sion)(Kendleretal.,2001).Itfollowsthatinapersonwithastrongdiathesisaprecipitatingminorstressor(whichwouldnotbeidenti-?edasa‘lifeevent’)mightbeeasilyoverlooked,whereasapersonwithaweakdiathesiswouldonlysuccumbtoaseverestressor(MonroeandHarkness,2005;HarknessandMonroe,2006).

Furthermore,therelationshipbetweenstressanddepres-sionchangesovertime.Depressionisarecurrentcondition,inwhicheachepisodeincreasestheprobabilityofafurtherepisode(Solomonetal.,2000;AmericanPsychiatricAssociation,1994).Indeed,oneofthemostpowerfulriskfactorsformajordepressionis

apre-existingchronicminordepression(Fogeletal.,2006).How-ever,therearemanystudiesdemonstratingthattheassociationbetweenseverestressanddepressiondecreaseswitheachsucces-siveepisode,sinceminorstressepisodesmaybynowbeasuf?cienttrigger(Dienesetal.,2006;Stroudetal.,2008;Morrisetal.,2010;Slavichetal.,2011).Forexample,astudyofover2000womenreportedthatoverthe?rstsixepisodesofdepressionthelikeli-hoodofanepisodeoccurringinanygivenmonthincreased15-fold,butatthesametime,theassociationwithastressfullifeeventdecreasedby75%(Kendleretal.,2000).Sooverthecourseofalifetimeofdepressiveepisodes,theonsetofdepressionbecomesincreasinglyautonomous.Thiseffectcanbedescribedbothneu-robiologically:anepisodeofdepressionsensitizes,or‘kindles’,thebraintorespondtoweakerandweakerprecipitants(Post,1992;Stroudetal.,2011),andpsychologically:thedepressedpersonreliesincreasinglyonnegativemodesofinformationprocessingthatcometobeactivatedbyincreasinglyminimalcues(Segaletal.,1996;Monroeetal.,2007).(Thesetwoaccountsfromdifferentstand-pointsareentirelycompatiblewithoneanother.)

Thus,therequirementforstrongandidenti?ablestressorstoprecipitateanepisodeofdepressiondecreasesasthedepressivediathesisincreases(e.g.severalpreviousepisodesofdepressionandahighgeneticrisk).However,inapersonwithaweakdepressivediathesis(e.g.nohistoryofdepressionandalowgeneticrisk)itisextremelylikelythatanepisodeofdepressionhasbeenprecipi-tatedbystress.

3.1.3.Implications

Thisanalysissuggeststheheuristicstrategythatwehaveadoptedintheremainderofthisreview.Weproposethatafullunderstandingoftheneurobiologyofdepressionimpliesasolutiontothreeseparateproblems:(1)themechanismsbywhichstressprecipitates?rstepisodesofdepressioninindividualswithaweakdepressivediathesis;(2)thekindlingprocessbywhichdepres-sionbecomesincreasinglyautonomousofstressoverrepeatedepisodes;and(3)thepsychologicalandneurobiological‘prekin-dling’processesbywhichsomeindividualsacquireastrongdepressivediathesisevenpriortotheir?rstepisodeofdepression.Wealsoproposethatthe?rstoftheseproblemscouldbeconsid-eredthemorefundamental,becauseasolutiontoitwouldprovideabasicmodelwithinwhichtodevelopsolutionstotheothertwoproblems:forexample,reportsofgeneticpolymorphismsthatcon-fervulnerabilitytodepressionareimpossibletointerpretwithoutamodelinwhichtounderstandwhatroleisplayedbythecellsinwhichthegeneproductisexpressed.Inthefollowingsections(Sections3.2–3.9)we?rstoutlinethemechanismsthatunderlystress-induced?rstepisodesofdepression:weareconcernedherewithunderstandingtheimpactofstressonparticularlyvulnera-blebrainregions,elucidatingtheneuralcircuitrythatisimpactedbytheseeffects,andrelatingtheseneurobiologicalchangestothecognitiveandbehaviouralfeaturesofdepression.Wesubsequentlyconsiderthe‘kindling’(Section3.10)and‘prekindling’(Section3.11)mechanismsthatincreasevulnerabilitytodepressionanddecreasetheroleofstress.

3.2.EffectsofstressontheHPAaxisandhippocampus

Themajorphysiologicalresponsetostressisanactivationofneuroendocrinesystems,mostnotably,thehypothalamus-pituitary-adrenal(HPA)axis.Inthissystem,corticotrophinreleasingfactor(orhormone:CRF/CRH)isreleasedfromthepara-ventricularnucleusofthehypothalamustostimulatethepituitaryglandtoproduceadrenocorticotrophichormone(ACTH),whichinturnstimulatesthereleaseofglucocorticoids(cortisolinhumansorcorticosteroneinrodents)fromtheadrenalcortexintothebloodcirculation,whichinteraliaexertnegativefeedbackeffectsonthe