Selective lentiviral-mediated suppression of microRNA124a in the hippocampus evokes antidepressants-like effects in rats

Selective

lentiviral-mediated

suppression

of

microRNA124a

in

the

hippocampus

evokes

antidepressants-like

effects

in

rats

Amine

Bahi

*

,

Vijay

Chandrasekar

,

Jean-Luc

Dreyer

a

DepartmentofAnatomy,TawamMedicalCampus,CMHS,UnitedArabEmiratesUniversity,AlAin, UnitedArabEmirates

b

InstituteofNeuropathology,UniversityHospitalofZurich,Schmelzbergstrasse12,CH-8091Zurich,Switzerland

c

DivisionofBiochemistry,DepartmentofMedicine,UniversityofFribourg,CH-1700Fribourg,Switzerland

KEYWORDS

BDNF;

Socialdefeatstress; Depression;

Lentiviralvector; microRNAs; miR124a

Summary Severallinesofevidencessuggestthatthebrain-derivedneutrophicfactor(BDNF)is implicatedinthepathophysiologyofdepression.However,themolecularmechanismsare not fullyunderstood.InthecurrentstudyweaimedtoinvestigatehowgeneticmodulationofBDNFin thehippocampususingmicroRNa124a(miR124a)-expressinglentiviralvectors(LV)mightaffect depression-like behaviorin adult rats. For this purpose, we assessed theexpression level of miR124aanditsdirecttargetBDNFinthehippocampusandthecortexafter21-daysexposureto socialdefeatstress.ResultsdemonstratedthatmiR124awasup-regulatedinthehippocampusbut notin thecortex.Incontrast, and asexpected,BDNFtranscripts weredown-regulated. In a different setof experiments, maleWistarratsreceivedbilateral intra-hippocampal or intra-corticalinfusionsofBDNF-andmiR124a-expressinglentiviralvectorsanddepression-likebehavior wasassessedafter21-dayssocialdefeatstressusingthenoveltysuppressedfeeding,thesucrose preference and the forced swim tests. The results indicated that miR124a overexpression exacerbated depression-like behavior.However, an anti-depressant like effect was observed whenBDNF ormiR124a-silencers (siR124a)wereinjected into thehippocampus.Importantly, whenexpressedintothecortex,LV-miR124a,LV-siR124aandLV-BDNFhadnoeffectondepression. OurfindingsindicatethathippocampalmiR124aanditsdirecttargetBDNFplayanimportantrole in depression-like behavior.Taken together,the current results reveal, for the first time, a potentialmolecularregulationofmiR124aonBDNF,andthepronouncedbehavioralconsequences ofthisregulationshedlightonthemechanismsunderlyingBDNFanti-depressantpotential. #2014ElsevierLtd.Allrightsreserved.

Abbreviations: BDNF,brain-derivedneurotrophicfactor;FST,forcedswimtest;Hipp,hippocampus;LV,lentiviralvector;miRNA,microRNA;

NSFT,noveltysuppressedfeedingtest;SPT,sucrosepreferencetest.

* Correspondingauthorat:DepartmentofAnatomy,CollegeofMedicine&HealthSciences,UnitedArabEmiratesUniversity,POBox17666,

AlAin,UnitedArabEmirates.Tel.:+97137137516;fax:+97137672033.

E-mailaddress:[email protected](A.Bahi).

Published in 3V\FKRQHXURHQGRFULQRORJ\± which should be cited to refer to this work.

1.

Introduction

Depressionisaprevalent,highlydebilitatingmentaldisorder affectingupto15%ofthepopulationatleastonceintheir lifetime, with huge costs for society. Neurobiological mechanismsofdepressionarestillnotwellknown,although there is consensus about interplay between genetic and environmentalfactors (Masiand Brovedani,2011). Antide-pressantmedicationsarefrequentlyusedindepression,but atleast50%ofpatientsarepoorresponders,evento more recentlydiscoveredmedications.Thereisaclearevidence foralteredendocrinefactorsandmetabolicdysregulationin mooddisorders(Hendrickxetal.,2005). However,clinical responseonlyoccursfollowingweekstomonthsoftreatment andonlychronictreatmentiseffective(MasiandBrovedani, 2011),suggestingthatactionsbeyondtherapidlyoccurring effectofenhancingmonoaminergicsystems,suchas adapta-tion of these systems, are responsible for the effects of antidepressants.

Recentstudieshaveshownthatcertainaspectsof depres-sion result from maladaptive stress-induced neuroplastic changes in specific neural circuits of the CNS (Nestler etal.,2002).Thesestudiesemphasizethatneuronal plas-ticityplaysanimportantroleintherecoveryfrom depres-sion and they indicate that an impairment of synaptic plasticity,particularlyinthehippocampus,maybeacore factorinthepathophysiologyofdepression.Theabnormal neuralplasticitymayberelatedtoalterationsinthelevelsof neurotrophic factors, namely brain-derived neurotrophic factor (BDNF), which play a central role in plasticity. As BDNFis repressed by stress,epigenetic regulation of the BDNFgenemayplayanimportantroleindepression(Masi andBrovedani,2011).Inturnasatreatment,antidepressant drugsandelectroconvulsiveshocktreatmentincreasethe expressionofBDNFanditsreceptorTrkB(Castrenand Ran-tamaki,2010).

Neurotrophinsconstituteafamilyofproteinsthatstimulate andcontrol neurogenesis during development and BDNF is amongthemostactive (Zigova etal.,1998).BDNF playsa keyrole on the central and peripheral nervous system to support the survival of existing neurons, and to induce neurogenesisandstimulatesynapticplasticityand synapto-genesis(Acheson etal.,1995). Itis mainly active in brain areasvitaltolearning,memory,andhigherthinking,including thehippocampus,thecortex,andbasal forebrain(Yamada andNabeshima,2003).Inaddition,BDNFregulatessynaptic plasticity in neuronal networks involved in several brain disordersincludingbipolardisorder,aswellasschizophrenia, obsessive—compulsivedisorder,Alzheimer’s disease, Hun-tington’sdisease,Rettsyndrome,anddementia(Schinder andPoo,2000).Furthermore,stressexposuredecreasesthe expressionofBDNF,and,ifexposureispersistent,thisleads toaneventualatrophyofthehippocampus,whichalsotakes placeinhumanssufferingfromchronicdepression.Inturn, regulationof BDNFmay reverse stress-induceddeficitsin structural and synaptic plasticity in the brain, yielding increased ability to cope with environmental challenges that may precipitate or exacerbate depressive episodes (CastrenandRantamaki,2010).LowBDNFanddepression aremost probablycausallyrelated,sinceantidepressants indeedincreaseBDNFsignalingandsynthesisinthe hippo-campus (Schmidt and Duman, 2010). Although it is now

firmly establishedthat BDNFsignaling playsanimportant roleinthemechanismofactionofantidepressantdrugs,yet theroleofBDNFinthepathophysiologyofdepressionisless clear(CastrenandRantamaki,2010).Furthermore,thereis strongevidencethat peripheral growthfactors,including BDNF,pro-inflammatorycytokines,endocrinefactors,and metabolic markers contribute to the pathophysiology of major depressive disorder and antidepressant response (Schmidt et al., 2011). As a matter of fact BDNF levels aredecreasedinthebloodofdepressed patientsand can be normalized with successful antidepressant treatment (Sen et al.,2008), althoughthe origin androle of serum BDNFis unclear(Karegeetal., 2005). Chronicperipheral administrationofBDNFproducesantidepressantand anxio-lyticbehavioralresponsesinanimalmodels,increasesthe survival rate of newborn neurons, and increases BDNF-mediated signaling in the adult hippocampus (Schmidt andDuman,2010),anevidencethatperipheral BDNFhas functionalactionswithinthebrainandonbehavior,andthat serumBDNFmaybearelevantbiomarkerfordepressionand treatmentresponse.

Neurotrophin BDNF acts via its receptor Tropomyosin-relatedkinaseBreceptor(TrkBorp75),promoting intracel-lular signalingthatis yieldingto protein synthesis (Soppet et al., 1991). The induction of protein synthesis by BDNF criticallycontributesto enduring modifications ofsynaptic function, buthowBDNF selectivelyaffectsonlya minority ofexpressedmRNAsispoorlyunderstood.Ithasbeen estab-lished thatBDNF rapidly elevatesDicer, increasing mature microRNAs (miRNA) levels and inducing RNA processing bodiesinneurons(Huangetal.,2012).

MiRNAsaresmallnon-codingRNAsofabout22-nucleotides inlengthregulatinggeneexpressionatpost-transcriptional level that act as key regulators of all cellular pathways (Treiberetal.,2012).Aftertranscription,thematuremiRNA is incorporated into a miRNA—protein complex, where it directlyinteractswith amemberoftheArgonauteprotein family.ThemiRNAguidessuchproteincomplexestopartial complementarytargetsites,whichare typicallylocatedin the30untranslatedregion(UTR)ofmRNAsleadingto inhibi-tion of gene expression.MiRNA activity and abundance is regulated on various levels ranging from transcriptionand processingtotargetsitebindingandmiRNAstability(Treiber etal.,2012).AsinglemiRNAmayspecificallyregulategene expressionofoverhundreddifferenttargets.Amongthem, miR124aisstronglyexpressedinthebrain;miR124aregulates adultneurogenesis(Caoetal.,2007),andpromotesneuronal differentiation(Makeyevetal.,2007) anditisinvolved in severalbraindisorders(Qietal.,2012).Inaddition,miR124a also strongly affects the expressionof several neurotrans-mitters,includingmainlyBDNF(BahiandDreyer,2013; Chan-drasekar and Dreyer, 2009, 2011). This prompted us to investigate thepotential roleof miR124ain the ethiology ofdepression.

Forthispurpose,inthisstudyweassessedtheexpression levelsofmiR124aanditsdirecttargetBDNFintheratbrain followingexposureto21-dayssocialdefeatstress.In addi-tion,ourmajorgoalinthepresentstudieswastoexamine the effects of lentiviral-mediated genetic modulation of BDNFandmiR124a inmaleWistarrats onperformancein threebehavioraltestsfollowingexposuretochronicsocial defeatstress.

2.

Materials

and

methods

2.1. Animals

ExperimentalmaleWistarrats(200g)weresingle-housed in standard plexiglas cages at a room temperature of approx. 228C with a 12h/12h light/dark cycle and allowed to adapt to this environment for a period of 7-daysbeforethe experimentsbegan. Allratswere bredin thelocalcentralanimalfacilityoftheCMHSandwerekept under standard laboratory conditions. Bedding was pro-ducedlocallyandautoclavedbeforeuseandratshadfree access to tap water and standard rodents chow diet obtained from the National Feed and Flour Production and Marketing Company LLC (Abu Dhabi, UAE). All the experimental procedures were approved by the institu-tional ResearchEthics Committee (Protocol NoA17-12).

2.2. TissuecollectionandquantificationofBDNF

andmiR124a aftersocialdefeatstress

Brains from decapitated control and stressed rats were rapidly removed and brain regions(whole cerebralcortex andhippocampus)weredissectedoutonice.Tissuesamples wereimmediatelyfrozeninTRizolandstoredat808Cuntil analysis. The micro-dissection procedure was performed accordingtorat stereotaxiccoordinates (Paxinosand Wat-son,1998).

Reactions were essentially carried out as described previously (Bahi and Dreyer, 2013; Chandrasekar and Dreyer,2009). Inbrief, total RNA from brain regionswas isolated using the TRizol according to the manufacturer instructions. Total RNA (5mg per sample) was reverse transcribed using Oligo-dT standard primers. For qPCR, 20ml of amplification mixture with SYBRGreen Kit was usedcontaining 3ml of cDNAand100nM primers. Primer sequences used were: GAPDH: 50-ATG ACT CTA CCC ACG GCA AG-30 and 50-CAT ACT CAG CAC CAG CAT CAC-30; miR124a: 50-TCC GTG TTC ACA GCG GAC-30 and 50-CAT TCA CCG CGT GCC TTA-30 and BDNF: 50-GGT TCG AGA GGT CTG ACG AC-30 and 50-CAA AGG CAC TTG ACT GCT GA-30(Bahietal.,2008a;BahiandDreyer,2013).ThePCR protocol used consisted of a 45-s denaturation at 948C followed by 45-sannealing and extensionat648C for 40 cycles.

2.3. Lentiviralconstructionsand production

Ourlentivirusderivedplasmid(pTK431),drivenbythe cyto-megalovirus (CMV) promoter, carried the transgenes for BDNF,miR124aorsiR124adescribedinourpreviousstudies (Bahietal.,2008a;BahiandDreyer,2013;Chandrasekarand Dreyer, 2009). The empty vector (LV-Mock) was used as negative control. All constructions were confirmed by sequencing. Viral vector generation was obtained by co-transfection of HEK293T cells by the calcium phosphate method on 10-cm plates with 20mg of pTK431, 15mg of pDNRF,and5mgofpMDG-VSV-G.Thesupernatantwas har-vested24 and72hafterinfection andviral particleswere purifiedbyultracentrifugation(Bahietal.,2005,2008b).

2.4. StereotaxicinjectionofLV-BDNF,LV-Mock,

LV-miR124aandLV-siR124a

For stereotaxic surgery, rats were anesthetized with a ketamine/xylazine mixture (100mg/kg and 10mg/kg respectively, i.p.) and installed in a stereotaxic frame. Using a precision Hamilton micro-syringe with a 26G needle, rats werebilaterally infused with1mlviral solu-tion using the following coordinates: hippocampus [1st injection: 4.8mm posterior to Bregma, 2.5mm lateral themedialsuture,3.5mmventraltotheskullsurface.2nd injection:4.8mmposteriortoBregma,5mmlateralthe medialsuture,6mm ventral totheskullsurface].Cortex [1stinjection:2.2mmanteriortoBregma,3mmlateral themedialsuture, 2mm ventraltotheskullsurface.2nd injection: 2.2mm anterior to Bregma, 4.5mm lateral the medial suture, 4mm ventral to the skull surface]. All viral-injections were performed according to the rat stereotaxiccoordinates(Paxinosand Watson,1998).Rats werelefttorecoverfor7daysbeforebeingexposedtothe socialdefeatstress procedure.After recovery andin the 1stsetofexperiments,weendedupwith19ratsthatwere injected into the hippocampus (LV-Mock n=9, LV-BDNF n=10) and 16 rats that were injected into the cortex (LV-Mockn=8,LV-BDNFn=8).Inthe2ndsetofexperiment we ended up with 24 rats that were injected into the hippocampus (LV-Mock n=7, LV-miR124a n=9, siR124a n=8) and 20 rats that were injected into the cortex (LV-Mockn=6,LV-miR124an=7,siR124an=7).

2.5. Social defeatstress

Thesocialdefeatstressmodelofdepressionwasbasedonthe resident-intruder paradigm (Korte et al., 1990). In brief, experimental-ratswereexposedtoa largedominantadult (400g)maleWistaraggressorratfor30mineachdayovera totalof 21 days. Aftercontact, the experimental-rat was separatedfromtheaggressorresident-ratusingacagethat wasdividedintotwopartsbyaperforateddivider.Thus,the experimental-rat andthe resident-rat were maintained in visualandsensorycontactforthenext24hwithoutphysical contact.Every4thdaytheexperimental-ratwasexposedtoa new aggressor. Resident aggressor rats were selected for theirlevelofaggressivenessbycountingthelatencytoattack anaiveWistarrat(maximum2mindelay).Topromotetheir aggressive behavior, the resident-rats were housed with females 10 days prior to the social defeat stress test to ensurethattheyacceptanddefendtheircageashomecage (territory).

2.6. Depression-likebehavioraltests

2.6.1. Noveltysuppressedfeeding(NSF)test

Thenoveltysuppressedfeedingprocedurewasperformedas describedpreviously (BahiandDreyer,2012).Inbrief, 24h priortotesting,ratsweredeprivedoffoodbutnotwater.On thetest day, eachrat was introducedintoa whiteplastic open arena (32cm32cm) covered with wood-chip bed-ding,andonwhichcenterseveralpelletsoffoodwereputon a circular paper.The latency to bite the food pellet was

measuredasthetestmainparameter.Immediatelyafterthe animalcommencedeating,theratwasimmediatelyremoved from the testing box and returned to its home cage with accesstofood.Therefore,thetotalpelletconsumptionwas measuredfor afurther 90min periodin order to evaluate homecagefoodintake.

2.6.2. Sucrosepreferencetest(SPT)

FortheSPT,atwo-bottlechoiceprocedurewasusedtotest fordifferences betweenthe groups for theirrelative pre-ference for sucrose over water. The rats were given free choicebetween2%sucrose-solutionandtapwaterfor24h. Oneweekbeforestarting,experimental-ratshadbeen habi-tuatedtotwo-bottlechoiceconditionsandthepositionofthe bottleswas switchedafter every24h to prevent anyside preferenceindrinkingbehavior.Nopreviousfoodorwater deprivation was applied before the test. Preference was measuredasfollows:(i)sucroseconsumption(ml);(ii)water consumption(ml);and(iii)totalliquid(sucrose+water).At the end of the test, the bottles were removed and the consumption was noted. The preference for sucrose was calculated as percentage of consumed sucrose-solution to

thetotalamountof liquiddrunk(Hennebelleetal.,2012; Zuritaetal.,1996).

2.6.3. Forcedswimtest(FST)

The FSTwas performed accordingto the methodoriginally reportedbyPorsoltandcolleagues(Porsoltetal.,1977)and thisprocedurewassuccessfullyusedinthelaboratoryinour previousstudies(BahiandDreyer,2012).Inbrief,glassbeakers werefilledwithtapwater(22—268C).Ratswereplacedinto thetestbeakerandwereunabletoescapeorrestbytouching thebottomofthebeaker.Thecylinderswereemptiedand cleanedbetweenrats.Sessionslasted6minandthefollowing behavioralparametersweremeasured:immobility,swimming andclimbing.Immobilitywasdefined astheabsenceofall motions except those required for keeping the rat’s head abovewater.Theswimmingbehaviorwasconsideredifarat wasactivelymakingswimmingmovementsthatcauseditto movewithinthecenterofthecylinder.Theclimbingbehavior wasrecordedifaratwasmakingforcefulthrashingmovements withitsforelimbsagainstthewallsofcylinder.Thedataare expressedasthemeantime(SEM)ofimmobility,swimming andclimbingwithinthe360-sobservationperiod.

Figure1 Effectofchronicsocialdefeaton(A)BDNFand(B)miR124aexpressionintherathippocampusandcortex.Expressionlevels werenormalizedagainstGAPDH inthecorresponding samples. Histogramsrepresent meanSEM,n=6. *Significantdifferences betweencontrolandstress(p_<0.05).#_{Significantdifferencesbetweenthehippocampusandthecortex(p<0.05)(two-wayANOVA,}

Bonferroni’sposthoctest),(C)effectoflentiviralinjectiononBDNFmRNAexpressioninthedorsalDG,ventralDGandtheCA1/CA3. ExpressionlevelswerenormalizedagainstGAPDHinthecorrespondingsamples.Histogramsrepresentmean_SEM,n=7.*Significant differencesbetweenLV-MockandLV-miR124a(p<0.05).**SignificantdifferencesbetweenLV-MockandLV-siR124a(p<0.001) (two-wayANOVA, Bonferroni’s post hoctest)and (D) schematic representationof the experimental procedures.The timeline shows sequenceanddurationof experimentalprotocolsoftheeffectofviralinjectiononchronicsocialdefeat-induceddepression-like behavior.

2.7. Statisticalanalysis

Forstatisticalcomparisons,thesoftwarepackageSPSS (ver-sion20.0)wasused.AlldatawereexpressedasmeansSEM. BDNF andmiR124amRNA expressionwas analyzed using a two-wayanalysisofvariance(ANOVA)withtreatment (con-trolvs.stress)andregion (hippocampusvs.cortex)as the betweensubjectfactors.Dataobtainedfromthebehavioral testswereanalyzedusingatwo-wayANOVAwithvirusand region asthe betweensubjectfactors. Posthocindividual meancomparisonswereperformedwiththeBonferroni’stest whenF valuesweresignificant.Thelevelofstatistical sig-nificancewassetatp<0.05atalltimes.

3.

Results

3.1. Effectsofsocialdefeatstressonbrain

expression ofBDNFandmiR124a

Fig.1showstheeffectsofchronicsocialdefeatstressonthe expression of BDNF in the hippocampus and the cortex detectedbyRT-PCR.AsshowninFig.1A,21-dayssocialdefeat stresscausedasignificantdownregulation(4folds)ofBDNF mRNAinthehippocampusbutnotinthecortex.Atwo-way ANOVA revealed a main effect of stress (F(1,20)=4.027,

p=0.058);butthemaineffectofbrain regionwasnot sig-nificant(F(1,20)=1.533,p=0.230).Nevertheless,the

interac-tion between stress and brain region was significant

(F(1,20)=10.926,p=0.004).Thesefindingsdemonstratedthat

3weeksofsocialdefeatstressinratsdown-regulated hippo-campal,butnotcortical,BDNFmRNAexpression.

Fig.1BshowsthemiR124aexpressionlevelsinthe inves-tigated brainregionshippocampus andcortex.All dataare meansSEMoftheparticularbrainspecimensfromthe con-trol(n=6)andsociallydefeated(n=6)groupsofrats.Results showthatmiR124aisexpressedineachoftheinvestigated brain regions, although at different levels. In fact, social defeat stress produced a significant increase (2 folds) of miR124ainthehippocampusbutnotinthecortex.Two-way ANOVAindicatedthattheindividual [stress:(F(1,20)=8.689,

p=0.008); brain region (F(1,20)=3.446, p=0.078)] and

interactive (F(1,20)=4.784, p=0.041) effects of stress and

brain regions were statistically significant. These results indicatedthat3-weeksofsocialdefeatstressinratsincreased miR124a expression into the hippocampus, but not into thecortex.

InordertoprovidemoreconvincingevidencethatBDNF wasdirectlyregulatedbymiR124A,ratswerestereotaxically injectedwithLV-Mock,LV-miR124aorLV-siR124ainthe hip-pocampusandBDNFmRNAwasquantifiedusingRT-PCR.The regionschosentoquantifyBDNFexpressionwerethedorsal andventraldentategyrus(DG)andtheCA1/CA3region.As depictedinFig.1C,two-wayANOVAanalysiswithvirusand sub-region as thebetween subject factors revealed a sig-nificanteffectofviralinjection(F(2,54)=28.992,p<0.0001)

but no effect of sub-region was detected (F(2,54)=1.653,

p=0.205). Nevertheless, theinteraction betweenthetwo factorswas significant(F(4,54)=7.297,p<0.001).Posthoc

evaluations revealed that LV-miR124a and LV-siR124a only affectedBDNFexpressioninthedorsalDGaswellasinthe CA1/CA3sub-regionsbutnotintheventralDG.

Based on these findings we decided to investigate the potential protective role of BDNF and miR124a-silencers overexpression into the hippocampus on social defeat stress-induceddepression-likebehaviorinrats.The experi-mentaltimelineisdepictedinFig.1D.

3.2. EffectsofBDNFoverexpressiononsocial

defeatstress-induceddepression

The novelty-induced suppression of feeding (NSF) test assessesdepression-andanxiety-likephenotypesby measur-ingthelatencyofratstoeatfoodinanovelanxiogenicopen fieldfollowingfooddeprivation(Bodnoffetal.,1989).Rats werestereotaxicallyinjectedintothehippocampusorinto thecortexandexposedtothesocialdefeatstress7dayslater asdescribedinSection2.Weshouldemphasizethatonerat from eachgroup was removed from the analysis as they escapedfromthe arena.Therefore the groupsizeswere: hippocampus (LV-Mock n=8, LV-BDNF n=10); cortex (LV-Mock n=8, LV-BDNF n=7). Using a two-way ANOVA test wefoundasignificanteffectofvirusforthelatencytobite afoodpellet at thecenterof thearena, withthe hippo-campusBDNF-injectedgrouphavingasignificantlyshorter latency to feed [main effect of virus: (F(1,29)=13.704,

p=0.001); main effect of brain region: (F(1,29)=11.558,

p=0.002)]. More importantly, the interaction between the twofactors was significant (F(1,29)=9.918,p=0.004)

(Fig.2A).InadditionandasdepictedinFig.2B,BDNFratsdid notdifferfromMock-injectedratsintheirfoodintakeinthe homecage during a 90-minperiod immediatelyfollowing the latency test [main effect of virus: (F(1,29)=0.082,

p=0.776); main effect of brain region: (F(1,29)=0.048,

p=0.829); interaction between virus and brain region:

(F(1,29)=0.254,p=0.618)].Finally,bothMock-and

BDNF-overexpressing groups did not differ in the amount of weightlost during the24h fooddeprivation period(data not shown).Thesefindings stronglysuggest that reduced latencytofeedwasdirectlyrelatedtoloweranxietylevelsin thehippocampus,butnotthecortex,of BDNF-overexpres-singrats.

Inorderprovideanappreciationoftheprecisemagnitude ofthestressorsaloneonthevariousbehavioraltestsandhow thevariouslentiviraltreatmentsaloneaffectedthesetests, weperformedafurtherexperimentusingnon-stressed con-trolgroupthatisgiventhevariouslentiviraltreatmentsin the hippocampus (not the cortex) and then subsequently tested(SupplementaryFig.1).Itshouldbeemphasizedthat thedatawith non-stress controls andstressanimals are a combinationofseparateexperiments,whichisfarfromthe optimaldesign.

Ratswere thenassessed in thesucrose preferencetest (SPT). This is a model of anhedonia, a core symptom of depression, in which rats fail to elicit pleasure from an activity that is normally enjoyable (Sofia and Knobloch, 1976).Thegroupsofratsusedwereasfollow:hippocampus (LV-Mockn=9, LV-BDNF n=10); cortex (LV-Mock n=8, LV-BDNFn=8).TheSPTscoresareshowninFig.2Candthese datafollowthesamepatternastheNSFtestdata.Infact,a two-way ANOVArevealed that lentiviral-mediated overex-pressionofBDNFinthehippocampus,butnotinthecortex, increasedsucrosepreference.Indeed,therewassignificant

main effect of brainregion (F(1,31)=8.897, p=0.006)and

viralinjection(F(1,31)=4.856,p=0.035).Moreimportantly,

theinteractionbetweenviral-injectionandthebrainregion was significant (F(1,31)=10.433, p=0.003). As a possibly

confounding variablefor anhedonia, totalfluid intake was alsotestedandresultsaredepictedinFig.2D.Thetwo-way ANOVAanalysis revealedthat there wasno main effect of viral injection (F(1,31)=1.170, p=0.288) nor brain region

(F(1,31)=0.169,p=0.863)ontotalfluidintake.Interestingly,

theinteractionbetweenviral-injectionandbrainregionwas notsignificant(F(1,31)=0.104,p=0.749).The experiments

usingthenon-stressedcontrolgroupsareshownin Supple-mentaryFig.2.Thedatawithnon-stresscontrolsandstress animalsareacombinationofseparateexperiments,whichis farfromtheoptimaldesign.

After completion of the SPT, rats were tested in the Forcedswimtest(FST)48-hlaterandresultsaredepicted in Fig. 3. The groups of rats used were as follow after removing 3rats becauseof diving: hippocampus(LV-Mock n=8,LV-BDNFn=9);cortex(LV-Mockn=7,LV-BDNFn=8). Two-wayANOVAanalysisrevealedthatBDNFoverexpression in the hippocampus, but not in the cortex, significantly decreasedthetimeofimmobility(2.5fold)[maineffect ofbrainregion: (F(1,28)=4.820,p=0.037),maineffectof

viral-injection(F(1,28)=3.849,p=0.060),themaineffectof

brainregionvirusinteraction(F(1,28)=6.147,p=0.019)]

(Fig.3A).Consequentlyandasexpected,wheninjectedinto thehippocampus,butnotintothecortex,BDNFsignificantly increasedthetimetheratsspentswimming[maineffectof brain region: (F(1,28)=4.217, p=0.049), main effect of

viral-injection(F(1,28)=2.847,p=0.103),the maineffect

oftheinteractionbetweentheviral-injectionandthebrain region (F(1,28)=7.607, p=0.010)] (Fig. 3B). The

experi-mentsusingthe non-stressedcontrol groupsareshownin SupplementaryFig.3.Thedatawithnon-stresscontrolsand stressanimalsareacombinationofseparateexperiments, whichisfarfromtheoptimaldesign.

Ourpresentdataindicatethatlentiviral-mediated over-expressionofBDNFintothehippocampus,butnotintothe cortex,inducesreductionofimmobilityandenhancementin theswimmingtimeintheFSTinratsfollowingchronicsocial defeatstress.

3.3. EffectsofmiR124aoverexpressionand

silencingonsocialdefeatstress-induced depression-likebehaviorinrats

To test the effects of miR124a on social defeat stress-induceddepressionlikebehaviorinWistarrats,both gain-and loss-of-function approaches were used. In these experiments,LV-siR124awasusedtoknockdownits expres-sionandalentivirusexpressingmiR124a(LV-miR124a)was usedtoup-regulateitsexpressioninthehippocampusorinto the cortex. Empty vectorwas usedas a negative control (LV-Mock).

Rats were first tested in the NSF test and results are depictedinFig. 4.Thegroupsofratsusedwereasfollow: hippocampus (LV-Mock n=7, LV-miR124a n=8, siR124a n=7); cortex (LV-Mock n=6, LV-miR124a n=7, siR124a n=7).Two-wayANOVAanalysishasshownthatthelatency to feed varied as a function of virus (F(2,38)=13.932,

p<0.001)butnot brainregion (F(1,38)=1.402,p=0.460).

Nevertheless, the interaction between viral-injection and the brain region was significant (F(2,38)=7.969, p=0.001)

(Fig.4A).Posthocevaluationsrevealedthat,comparedto

Figure2 Effectoflentiviral-mediatedBDNFoverexpressionon chronic social defeat-induced depression-like behavior in the noveltysuppressedfeeding (NSF)and sucrosepreferencetest (SPT).(A)Eatingdelayasmeasuredbythelatencytobitefoodin anovelanxiogenicenvironmentfollowing24-hfooddeprivation, (B)foodintakeofratsinthehomecageduringa90minperiod followingNSFlatencytest,(C)sucrosepreferencewas calculat-ed by the following formula: preference=(sucrose solution intake/totalfluidintake)100, and (D)total fluidintake su-crose intake+water intake. *Significant differences between Mock and BDNF (p<0.05). #Significant differences between thehippocampusand thecortex (p_<0.05)(two-wayANOVA, Bonferroni’s post hoc test). Hippocampus (LV-Mock n=9, LV-BDNFn=10);cortex(LV-Mockn=8,LV-BDNFn=8).

Figure3 Effectoflentiviral-mediatedBDNFoverexpressionon chronic social defeat-induced depression-like behavior in the forced-swimtest(FST).(A)Immobilityand(B)swimming. *Sig-nificantdifferencesbetweenMockandBDNF(p<0.05).# Signif-icant differences between the hippocampus and the cortex (p<0.05)(two-wayANOVA,Bonferroni’sposthoctest). Hippo-campus(LV-Mockn=9,LV-BDNFn=10);cortex(LV-Mockn=8, LV-BDNFn=8).

controls(LV-Mock),ratsinjectedwithmiR124ashowedaca. 1.5-fold increased latency to approach the food pellet (p=0.028). More importantly and as expected, silencing miR124aexpressionintothehippocampus,butnotintothe cortex,usingsiR124adrasticallyreducedthelatencytofeed (p=0.034 and p<0.001 vs. Mock and miR124a respec-tively). In addition and as depicted in Fig. 4B, the three-groupsofratsdidnotdifferintheirfoodintakeinthehome cage during a 90min period immediately following the latencytest [maineffectof brain region:(F(1,38)=0.243,

p=0.625);main effectof viral-injection: (F(2,38)=0.179,

p=0.837); interaction between virus and brain region:

(F(2,38)=0.126, p=0.882)].Finally,both

miR124a-overex-pressingand-silencinggroupsofratsdidnotdifferin the amount of weight lost during the 24h food deprivation period (datanot shown).Theexperiments usingthe non-stressedcontrolgroupsareshowninSupplementaryFig.4. Thedatawithnon-stresscontrolsandstressanimalsarea

combinationofseparateexperiments,whichisfarfromthe optimaldesign.

These findings strongly suggest that decreased and increased latency to feed were directly related to lower andhigheranxietylevels insiR124a- and miR124a-overex-pressingratsrespectively.

AftercompletionoftheNSFtest,thesameratswereassed inthesucrose preferencetest(SPT)usingthemethodology describedinSection2aboveandresultsaredepictedinFig.4C andD.Thegroupsofratsusedwereasfollow:hippocampus (LV-Mockn=7,LV-miR124an=9,siR124a n=8);cortex (LV-Mockn=6,LV-miR124an=7,siR124an=7).Two-wayANOVA analysisrevealedthattheaveragesucrosepreferenceofrats calculated from measures taken in Mock, siRNA124a and miR124a-injected groups differed significantly from each other.InfactandasshowninFig.4C,therewasasignificant maineffectofviral-injection(F(2,38)=30.598,p<0.001)as

well as brain region (F(1,38)=9.751, p=0.003) on sucrose

preference.Interestingly, theinteraction between thetwo factorwashighlysignificant(F(2,38)=43.635,p<0.001).Post

hocevaluationsindicatedthatcomparedtoMock-, miR124a-expressingratsshowed3foldsdecreasedsucrosepreference (p<0.001). In contrast, siR124a-injected rats displayed higherpreferenceforsucrose(1.3fold)comparedtoMock controlanimals(p=0.047).Duringthesametest,ratswere then assessedfor their total fluid intake and the two-way ANOVAanalysisrevealedthatnosignificantdifferenceswere observed.AsdepictedinFig.4D,Mock-,siR124a-and miR124a-injectedratsconsumedthesameamountoffluidduringthe SPT.Indeed,therewasnomaineffectofvirus(F(2,38)=0.142,

p=0.868)andnosignificantmaineffectofthebrainregion

(F(1,38)=0.018, p=0.894). More importantly, the virus

-brainregion interactionwasnotsignificant(F(2,38)=0.284,

p=0.754). Theexperiments using thenon-stressed control groups are shown in Supplementary Fig. 5. The data with non-stresscontrols andstress animalsareacombinationof separateexperiments,whichisfarfromtheoptimaldesign.

Figure4 Effectoflentiviral-mediatedmiR124amodulationon chronic social defeat-induced depression-like behaviorin the noveltysuppressedfeeding(NSF) andsucrosepreference test (SPT).(A)Eatingdelayasmeasuredbythelatencytobitefoodin anovelanxiogenicenvironmentfollowing24-hfooddeprivation, (B)foodintakeofratsinthehomecageduringa90minperiod followingNSFlatencytest,(C)sucrosepreferencewas calculat-ed by the following formula: preference=(sucrose solution intake/total fluid intake)100. *Significant differences be-tween miR124a or siR124a and Mock (p<0.05). #_Significant

differences between the hippocampus and the cortex (p<0.05) (two-way ANOVA, Bonferroni’s posthoc test), and (D)totalfluidintakesucroseintake+waterintake.*Significant differencesbetweenmiR124aorsiR124aandMock(p<0.05).

#_{Significantdifferencesbetweenthehippocampusandthe}

cor-tex (p<0.05) (two-way ANOVA, Bonferroni’s post hoc test). Hippocampus(LV-Mockn=7,LV-miR124a n=9,siR124a n=8); cortex(LV-Mockn=6,LV-miR124an=7,siR124an=7).

Figure5 Effectoflentiviral-mediatedmiR124amodulationon chronic social defeat-induced depression-like behavior in the forced-swimtest(FST).(A)Immobilityand(B)swimming. *Sig-nificant differences between miR124a or siR124a and Mock (p<0.05).#Significantdifferences between thehippocampus andthe cortex(p<0.05)(two-wayANOVA, Bonferroni’spost hoc test). Hippocampus (LV-Mock n=7, LV-miR124a n=9, siR124an=8);cortex(LV-Mockn=6,LV-miR124an=7,siR124a n=7).

AftercompletionoftheSPT,ratsweretestedintheFST usingthese groupsofanimals:hippocampus(LV-Mockn=6, LV-miR124a n=9, siR124a n=8); cortex (LV-Mock n=6, LV-miR124an=7,siR124an=7).AsdepictedinFig.5, mod-ulation of miR124a expression into the hippocampus profoundly affectedFST behavior.In fact,the duration of immobility was affected following viral-injection

(F(2,38)=19.285,p<0.001)but therewasno effectofthe

brainregion (F(1,38)=0.087,p=0. 770).Nevertheless, the

viralinjectionbrainregion interactionwashighly

signifi-cant(F(2,38)=24.208,p<0.001)(Fig.5A).Posthoc

evalua-tionindicatedthatmiR124a-overexpressingratswere1.7 foldmoreimmobilethanMockcontrolanimals(p=0.012).In contrastandasexpected,siR124a-expressingratswere3 foldslessimmobilethancontrols(p=0.005).Inadditionand asshowninFig.5B,thesamepatternwasalsoobservedwhen swimmingbehaviorwasassessedinthesegroups.Thus,the two-wayANOVAanalysisindicatedasignificantmaineffectof viral-injection(F(2,38)=16.586,p<0.001)andasignificant

interaction between the brain region and viral-injection

(F(2,38)=19.466,p<0.001).Posthoc evaluationsrevealed

thatcompared to Mockcontrols, miR124a spentless time swimming (p=0.071). However, siR124a-expressing rats weremoreactive intheir swimming behavior(p=0.003). Theexperimentsusing thenon-stressedcontrolgroupsare shown in Supplementary Fig. 6. The data with non-stress controlsand stressanimals are a combination ofseparate experiments,whichisfarfromtheoptimaldesign.

4.

Discussion

Thisisthefirststudytodemonstratethatexposuretochronic socialdefeat stress can inducemiR124a expressionin the brain hippocampus, a key brain region associated with depression-likebehaviorinrats.Inadditionandasexpected, themRNAexpressionlevelofBDNF,whichisadirecttargetof miR124a, was reduced following stress exposure. More importantly,weshowedthatviralmediatedoverexpression ofmiR124ainthehippocampus,butnotinthecortex,was pro-depressant.Incontrast,BDNForsiR124aoverexpression reduceddepression-likebehaviorinrats.Theseresults indi-cate that miR124a is a critical regulator for BDNF in rat hippocampusandwehypothesizeconsequentlythatmiR124a mightparticipateinthesocialdefeatstress-induced depres-sive-likebehavior through the regulation of BDNF genetic expression.

NeurotrophicfactorssuchasBDNFwereshowntoinduce changesinhippocampalneurogenesisandmediatethe neu-ropathophysiologyofdepression(Castrenetal.,2007;D’Sa andDuman,2002;DumanandMonteggia,2006).Tofurther clarifytheneurochemicalimpactofchronicstresson hippo-campalneurotrophic factors expression,we studiedmRNA levelsoftheBDNFafter21-dayssocialdefeatstress.Results haveshownthat,afterthree weeks, stressinduceda pro-nounceddecreaseinBDNFexpressioninthehippocampusbut notin thecortex confirmingprevious reports whichfound thatchronicrestraintorsocialdefeatstresscaused down-regulationofBDNFinthesamebrainregion(Haenischetal., 2009). In addition, our findings are in line with those of Pizarroandco-workerswhofoundusinginsituhybridization, that a brief (10min) exposure to intense social stress

significantlydecreasedBDNFmRNAexpressioninthe hippo-campus andamygdala 24h after socialdefeat when com-paredtocontrol(naive)mice(Pizarroetal.,2004).Also,our findingsagreewiththoseofAlfonsoandcolleagues,whohad shownthatwhenmiceweresubjectedtorepeatedrestraint stress,hippocampalBDNFmessengersweredown-regulated asmeasuredbyRT-PCR(Alfonsoetal.,2006).More impor-tantly,chronicsocialdefeat stresshasalsobeen shownto decreaseexpressionofhippocampalBDNF(Komatsuetal., 2011).Finally,itshouldbenotifiedthattheworkperformed inNestlerlabindicatedthatdetailedmRNAanalysisofwhole mouse hippocampus revealed that chronic defeat stress inducedanapproximatelythreefoldreductioninBDNFmRNA levels(Tsankovaetal.,2006).Therefore,wecould demon-stratethatchronicsocialdefeatstresscausedbrainregion specificalterationsintheexpressionofBDNFsuggestingthat thesechangesmayhaveimplicationsinbrainplasticityand behavioralchangesfollowingsocialstress.

InanattempttoclarifytheimpactofmicroRNAonBDNF messengers,we assessedtheexpressionof miR124ainthe hippocampusfollowing chronicsocial defeatstress. Micro-RNAsaresmallnon-codingRNAsinvolvedintheregulationof geneexpressionandproteintranslationandwehypothesize that miRNAs maybe involved in theregulation of chronic stress-relatedgenesingeneralanddepressioninparticular. ResultshaveshownthatreducedBDNFlevels were accom-paniedwithanincreaseofmiR124aexpressionafter21-days socialdefeat.Tothebestofourknowledge,wearethefirst who could demonstrate a negative correlation between levels ofBDNFandtheexpressionofmiR124a. Indeed, we haveshowninpreviousreportthatBDNFisadirecttargetof miR124a(BahiandDreyer,2013;ChandrasekarandDreyer, 2009).OurfindingsareinlinewiththoseofO’Connorand co-workerswhofoundusingmicroarrayandRT-PCR,that early-lifestressaffectedtheexpressionofmultiplehippocampal miRNAs (O’Connor et al.,2013). In fact, using a profiling approachithasbeenreportedthatearly-lifestressaffected the expression of multiple hippocampal miRNAs such as miR451 (O’Connor et al., 2013). When investigating the impactofsocialisolationongenesandassociated miRNAs, a recent study reported that the isolated rats exhibited higher levels of miR29 and miR203 (Yang et al., 2013). Importantly, the expression levels of vascular endothelial growthfactor Awhich isa directtarget ofthese miRNAs, wasstronglyandspecificallysuppressed(Yangetal.,2013). Inourpresentstudywedemonstrateforthefirsttimethat chronicsocialdefeatincreasesmiR124aexpression, suggest-ing a potential role for depression-associated gene and miRNAinteractionsviamodulationofBDNFexpression.

Social stress-induced alterations of gene expression in limbic areas are involved in long-term neuroadaptations andbehavioralimpairments (Martinez etal.,2002). Inter-estingly,anumberofreportslinkedBDNFneurotransmission intheneuropathophysiologyofpsychiatricdisordersin gen-eral (Balaratnasingam and Janca, 2012) anddepression in particular(DumanandMonteggia,2006).Unfortunately,the molecular mechanism by which BDNF may contribute to the pro-depressant effect of social stress is still not well understood.Inthecurrentstudywehavedemonstratedthat lentiviral-mediated overexpression of BDNF in the hippo-campus,butnotintothecortex,resultsinananti-depressant likebehavior.Thecurrentbehavioraldataareconsistentwith

resultsshowingthatinmaleSprague-Dawleyrats,abilateral infusionofBDNF(0.25and1.0mg)intothedentategyrusof thehippocampusincreasestheswimmingbehaviorintheFST and, therefore produces antidepressant-like phenotype (Shirayamaetal.,2002).Inaddition,Govindarajanand co-workers reported that genetically modified mice over-expressing BDNF in excitatory neurons of the forebrain, includingthehippocampus,displayedimprovedperformance inthePorsoltFST(Govindarajan etal.,2006).Incontrast, heterozygous(+/)BDNFmicewereresistanttotheeffects ofantidepressantsintheFST,indicatingthat antidepressant-induced behavioraleffects require regular BDNF signaling (Lindholmetal.,2012;Saarelainenetal.,2003).Inaddition, ithasbeenshownusingthesamemicethat,whencombined with a stress exposure, heterozygous depletion of BDNF resulted in a depressive phenotype in the FST (Duman etal.,2007).

Moreover, we found that hippocampalmiR124a overex-pression exacerbated stress-induced depression-like beha-vior. In contrast, inhibition of miR124a signaling in the hippocampususingalentiviral-mediatedsilencerexpression ‘‘rescued’’theanti-depressanteffectseenin BDNF-overex-pressing animals.Therefore, wehypothesize thatmiR124a playsapivotalroleinstress-relatedemotions,byregulating theinhibitoryeffectsofsocialdefeatstressonhippocampal BDNFexpression.Whatevertheunderlyingmechanisms,our datademonstratethatsiR124aandmiR124aexertopposite effectsonhippocampalBDNFmessengerlevels,andsuggest that the balance between these two factors may play a crucialroleindeterminingvulnerabilitytostress pro-depres-sant like effect. Taken together, we propose that social defeatstressexposureincreasestheexpressionofmiR124a andconsequentlydecreasestheexpression ofBDNFin the hippocampus.Therefore,theseresultssuggestthatmiR124a silencinghasanantidepressant-likeeffect,atleastinpartby supportingcellproliferationthroughtheBDNFsignaling path-way.

5.

Conclusion

Our study demonstratesthat chronicsocialdefeat impairs depression-likebehaviorinlaboratory animals.Theunique expressionpatternofgene—miRNA interactionsinthe hip-pocampus of socially defeated mice represents a novel aspectofthecomplexregulatorynetworkinvolvedin emo-tionalillnesses.Theseresultssuggestthatoverexpressionof miR124acontributestochronicsocialstressinduced depres-sion,partially bysuppressing BDNFexpression.Considering the important role of miR124a in emotions-related brain areas,theseresultssuggestaputativemechanismforchronic social defeat-induced behavioral impairments through the modulationofBDNF.TheinhibitionofBDNFidentifiedinthis studyandothers(BahiandDreyer,2013;Chandrasekarand Dreyer,2009)mayprovideatargetforfuture therapeutics thataredirected totacklepsychiatricdisordersingeneral anddepressioninparticular.

Role

of

the

funding

source

ABisreceivingfundsfromtheUnitedArabEmirates Univer-sity (No. NP/12/04,and NP/13/05)andfrom the National

Research Foundation(No. 31M082). JLD was supportedby SwissNationalScienceFoundation(Grantsnos.3100-059350, 3100AO-100686, and 31003A-116492). The funders had no roleinstudydesign,datacollectionandanalysis,decisionto publish,orpreparationofthemanuscript.

Conflict

of

interest

statement

Acknowledgements

TheauthorswouldliketoacknowledgeMrs.Christine Defor-el-Poncet forthe assistance with viralvectors cloning and preparation.We aregrateful alsoto Mr.MohamedElwasila andMr.MohamedShafiullahfortheirtechnicalassistance.

Appendix

A.

Supplementary

data

Supplementary material related to this article can be found, in the online version

References

Acheson,A.,Conover,J.C.,Fandl,J.P.,DeChiara,T.M.,Russell,M.,

Thadani,A., Squinto, S.P., Yancopoulos, G.D., Lindsay, R.M.,

1995.ABDNFautocrineloopinadultsensoryneuronsprevents

celldeath.Nature374,450—453.

Alfonso,J.,Frick,L.R.,Silberman,D.M.,Palumbo,M.L., Genaro,

A.M.,Frasch, A.C.,2006.Regulationofhippocampal gene

ex-pressionisconservedintwospeciessubjectedtodifferent stress-orsandantidepressanttreatments.Biol.Psychiatry59,244—251.

Bahi,A.,Boyer,F.,Bussard,G.,Dreyer,J.L.,2005.Silencing

dopa-mineD3-receptorsinthenucleusaccumbensshellinvivoinduces changesincocaine-inducedhyperlocomotion.Eur.J.Neurosci. 21,3415—3426.

Bahi,A.,Boyer,F.,Chandrasekar, V.,Dreyer,J.L.,2008a. Roleof

accumbensBDNFandTrkBincocaine-inducedpsychomotor sen-sitization,conditioned-placepreference,andreinstatementin rats.Psychopharmacology199,169—182.

Bahi,A.,Dreyer,J.L.,2012.Hippocampus-specificdeletionoftissue

plasminogenactivatortPAinadultmiceimpairsdepression-and anxiety-likebehaviors.Eur.Neuropsychopharmacol.22,672—682.

Bahi,A.,Dreyer,J.L.,2013.StriatalmodulationofBDNFexpression

usingmicroRNA124a-expressinglentiviralvectorsimpairs etha-nol-inducedconditioned-placepreferenceandvoluntaryalcohol consumption.Eur.J.Neurosci.38,2328—2337.

Bahi,A.,Kusnecov,A.W.,Dreyer,J.L.,2008b.Effectsof

urokinase-typeplasminogenactivator intheacquisition, expressionand reinstatementofcocaine-inducedconditioned-placepreference. Behav.BrainRes.191,17—25.

Balaratnasingam,S.,Janca,A., 2012.Brain derivedneurotrophic

factor:anovelneurotrophininvolvedinpsychiatricand neuro-logicaldisorders.Pharmacol.Ther.134,116—124.

Bodnoff,S.R.,Suranyi-Cadotte,B.,Quirion,R.,Meaney,M.J.,1989.

Acomparisonoftheeffectsofdiazepamversusseveraltypical andatypicalanti-depressantdrugsinananimalmodelofanxiety. Psychopharmacology97,277—279.

Cao,X.,Pfaff,S.L.,Gage,F.H.,2007.AfunctionalstudyofmiR-124in

thedevelopingneuraltube.GenesDev.21,531—536.

Castren,E.,Rantamaki,T.,2010.TheroleofBDNFanditsreceptors

indepression and antidepressant drugaction: reactivation of developmentalplasticity.Dev.Neurobiol.70,289—297.

Castren,E.,Voikar, V.,Rantamaki,T.,2007. Roleofneurotrophic factorsindepression.Curr.Opin.Pharmacol.7,18—21.

Chandrasekar,V.,Dreyer,J.L.,2009.microRNAsmiR-124,let-7dand

miR-181a regulate cocaine-induced plasticity. Mol.Cell. Neu-rosci.42,350—362.

Chandrasekar,V.,Dreyer,J.L.,2011.RegulationofMiR-124,Let-7d,

andMiR-181aintheaccumbensaffectstheexpression, extinc-tion, and reinstatement ofcocaine-inducedconditionedplace preference.Neuropsychopharmacology36,1149—1164.

D’Sa,C.,Duman,R.S.,2002.Antidepressantsandneuroplasticity.

BipolarDisord.4,183—194.

Duman,C.H., Schlesinger,L., Kodama,M., Russell,D.S., Duman,

R.S.,2007.AroleforMAPkinasesignalinginbehavioralmodelsof

depression and antidepressant treatment.Biol. Psychiatry61, 661—670.

Duman,R.S.,Monteggia,L.M.,2006.Aneurotrophicmodelfor

stress-relatedmooddisorders.Biol.Psychiatry59,1116—1127.

Govindarajan,A.,Rao,B.S.,Nair,D.,Trinh,M.,Mawjee,N.,Tonegawa,

S., Chattarji, S., 2006. Transgenic brain-derived neurotrophic

factor expression causes both anxiogenic and antidepressant effects.Proc.Natl.Acad.Sci.U.S.A.103,13208—13213.

Haenisch, B., Bilkei-Gorzo, A., Caron, M.G., Bonisch, H., 2009.

Knockoutofthenorepinephrinetransporterand pharmacologi-callydiverseantidepressantspreventbehavioralandbrain neu-rotrophinalterationsintwochronicstressmodelsofdepression. J.Neurochem.111,403—416.

Hendrickx,H.,McEwen,B.S.,Ouderaa,F.,2005.Metabolism,mood

andcognitioninaging:theimportanceoflifestyleanddietary intervention.Neurobiol.Aging26(Suppl.1) 1—5.

Hennebelle,M.,Balasse,L.,Latour,A.,Champeil-Potokar,G.,Denis,

S.,Lavialle,M.,Gisquet-Verrier,P.,Denis,I.,Vancassel,S.,2012.

Influenceofomega-3fattyacidstatusonthewayratsadaptto chronicrestraintstress.PLoSONE7,e42142.

Huang,Y.W.,Ruiz,C.R.,Eyler,E.C.,Lin,K.,Meffert,M.K.,2012.Dual

regulation ofmiRNAbiogenesis generates targetspecificityin neurotrophin-inducedproteinsynthesis.Cell148,933—946.

Karege,F.,Bondolfi,G.,Gervasoni,N., Schwald,M.,Aubry,J.M.,

Bertschy,G.,2005.Lowbrain-derivedneurotrophicfactor(BDNF)

levels in serum of depressed patients probably results from loweredplateletBDNFreleaseunrelatedtoplateletreactivity. Biol.Psychiatry57,1068—1072.

Komatsu,H.,Ohara,A.,Sasaki,K.,Abe,H.,Hattori,H.,Hall,F.S.,

Uhl, G.R.,Sora,I.,2011.Decreased responsetosocialdefeat

stress in mu-opioid-receptor knockout mice. Pharmacol. Biochem.Behav.99,676—682.

Korte,S.M.,Smit,J.,Bouws,G.A.,Koolhaas,J.M.,Bohus,B.,1990.

Behavioralandneuroendocrineresponsetopsychosocialstressin malerats:theeffectsofthe5-HT1Aagonistipsapirone.Horm. Behav.24,554—567.

Lindholm,J.S.,Autio,H.,Vesa,L.,Antila,H.,Lindemann,L.,Hoener,

M.C.,Skolnick,P.,Rantamaki,T.,Castren,E.,2012.The

antide-pressant-likeeffectsofglutamatergicdrugsketamineandAMPA receptor potentiator LY 451646 are preserved in bdnf(+)/() heterozygousnullmice.Neuropharmacology62,391—397.

Makeyev,E.V.,Zhang,J.,Carrasco,M.A., Maniatis,T.,2007.The

MicroRNAmiR-124promotesneuronaldifferentiationby trigger-ing brain-specificalternative pre-mRNA splicing.Mol.Cell27, 435—448.

Martinez,M.,Calvo-Torrent,A.,Herbert,J.,2002.Mappingbrain

response to social stress in rodents with c-fos expression: a review.Stress5,3—13.

Masi,G.,Brovedani,P.,2011.Thehippocampus,neurotrophicfactors

anddepression:possibleimplicationsforthepharmacotherapyof depression.CNSDrugs25,913—931.

Nestler, E.J., Barrot, M., DiLeone, R.J., Eisch, A.J., Gold, S.J.,

Monteggia,L.M.,2002.Neurobiologyofdepression.Neuron34,

13—25.

O’Connor, R.M., Grenham, S., Dinan, T.G., Cryan, J.F., 2013.

microRNAsasnovelantidepressanttargets:convergingeffects of ketamineand electroconvulsiveshock therapy in the rat hippocampus.Int.J.Neuropsychopharmacol.16,1885—1892.

Paxinos,G.,Watson,C.,1998.TheRatBraininStereotaxic

Coordi-nates.AcademicPress,SanDiego.

Pizarro,J.M.,Lumley,L.A.,Medina,W.,Robison,C.L.,Chang,W.E.,

Alagappan,A.,Bah,M.J.,Dawood,M.Y.,Shah,J.D.,Mark,B.,

Kendall,N.,Smith,M.A.,Saviolakis,G.A.,Meyerhoff,J.L.,2004.

Acute social defeat reduces neurotrophin expression inbrain corticalandsubcorticalareasinmice.BrainRes.1025,10—20.

Porsolt,R.D.,Bertin,A.,Jalfre,M.,1977.Behavioraldespairinmice:

aprimaryscreeningtestforantidepressants.Arch.Int. Pharma-codyn.Ther.229,327—336.

Qi,L.,Hu,Y.,Zhan,Y.,Wang,J.,Wang,B.B.,Xia,H.F.,Ma,X.,2012.

ASNPsiteinpri-miR-124changesmaturemiR-124expressionbut nocontributiontoAlzheimer’sdiseaseinaMongolianpopulation. Neurosci.Lett.515,1—6.

Saarelainen,T.,Hendolin,P.,Lucas,G.,Koponen,E.,Sairanen,M.,

MacDonald,E.,Agerman,K.,Haapasalo,A.,Nawa,H.,Aloyz,

R.,Ernfors,P.,Castren,E.,2003.ActivationoftheTrkB

neuro-trophin receptor is induced by antidepressant drugs and is required for antidepressant-induced behavioral effects. J. Neurosci.23,349—357.

Schinder, A.F., Poo, M., 2000. The neurotrophin hypothesis for

synapticplasticity.TrendsNeurosci.23,639—645.

Schmidt,H.D.,Duman,R.S.,2010.PeripheralBDNFproduces

anti-depressant-likeeffectsincellularandbehavioralmodels. Neu-ropsychopharmacology35,2378—2391.

Schmidt, H.D., Shelton, R.C., Duman, R.S., 2011. Functional

biomarkersofdepression:diagnosis,treatment,and pathophysi-ology.Neuropsychopharmacology36,2375—2394.

Sen,S.,Duman,R.,Sanacora,G.,2008.Serumbrain-derived

neuro-trophic factor, depression, and antidepressant medications: meta-analysesandimplications.Biol.Psychiatry64,527—532.

Shirayama,Y.,Chen,A.C.,Nakagawa,S.,Russell,D.S.,Duman,R.S.,

2002.Brain-derivedneurotrophicfactorproducesantidepressant

effects in behavioralmodels of depression. J. Neurosci. 22, 3251—3261.

Sofia,R.D.,Knobloch,L.C.,1976.Comparativeeffectsofvarious

naturally occurring cannabinoids on food, sucrose and water consumption by rats. Pharmacol. Biochem. Behav. 4, 591—599.

Soppet,D.,Escandon,E.,Maragos,J.,Middlemas,D.S.,Reid,S.W.,

Blair,J.,Burton,L.E.,Stanton,B.R.,Kaplan,D.R.,Hunter,T.,

Nikolics,K.,Parada,L.F.,1991.Theneurotrophicfactors

brain-derivedneurotrophicfactorandneurotrophin-3areligandsfor thetrkBtyrosinekinasereceptor.Cell65,895—903.

Treiber,T.,Treiber,N.,Meister,G.,2012.RegulationofmicroRNA

biogenesisandfunction.Thromb.Haemost.107,605—610.

Tsankova,N.M., Berton,O., Renthal,W.,Kumar, A., Neve,R.L.,

Nestler,E.J.,2006.Sustainedhippocampalchromatinregulation

inamousemodelofdepressionandantidepressantaction.Nat. Neurosci.9,519—525.

Yamada,K.,Nabeshima,T.,2003.Brain-derivedneurotrophic

fac-tor/TrkBsignalinginmemoryprocesses.J.Pharmacol.Sci.91, 267—270.

Yang,L.,Engeland,C.G.,Cheng,B.,2013.Socialisolationimpairs

oralpalatalwoundhealinginSprague-Dawleyrats:arolefor miR-29andmiR-203viaVEGFsuppression.PLoSONE8,e72359.

Zigova,T.,Pencea,V.,Wiegand,S.J.,Luskin,M.B.,1998.

Intraven-tricularadministrationofBDNFincreasesthenumberofnewly generatedneuronsintheadultolfactorybulb.Mol.Cell. Neu-rosci.11,234—245.

Zurita,A.,Murua,S.,Molina,V.,1996.Anendogenousopiate

mech-anismseemstobeinvolvedinstress-inducedanhedonia.Eur.J. Pharmacol.299,1—7.