Learning and memory disabilities in IUGR babies: Functional and molecular analysis in a rat model

Brain and Behavior. 2017;7:e00631.   | 1 of 11 https://doi.org/10.1002/brb3.631 wileyonlinelibrary.com/journal/brb3 Received:19March2016 | Revised:22July2016 | Accepted:30November2016 DOI:10.1002/brb3.631 ORIGINAL RESEARCH Learning and memory disabilities in IUGR babies: Functional and molecular analysis in a rat model Marta Camprubí Camprubí1 | Rafel Balada Caballé2 | Juan A. Ortega Cano2 | Maria de los Angeles Ortega de la Torre3 | Cristina Duran Fernández-Feijoo4 | Montserrat Girabent-Farrés5 | Josep Figueras-Aloy1 | Xavier Krauel1 | Soledad Alcántara2 1NeonatologyService,SantJoandeDéu,BCNatal,HospitalSantJoandeDéuiClínic,UniversityofBarcelona,Barcelona,Spain 2DepartmentofPathologyandExperimentalTherapeutics,SchoolofMedicine,UniversityofBarcelona,Barcelona,Spain 3DepartmentofMolecularBiologyandBiochemicalEngineering,UniversityPablodeOlavide–CABIMER–CIBERDEM,Sevilla,Spain 4PediatricsService,HospitalÁlvaroCunqueiro,Vigo,Spain 5UniversitatInternacionaldeCatalunya,SantCugatdelVallès,Barcelona,Spain ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium, providedtheoriginalworkisproperlycited. ©2017TheAuthors.Brain and BehaviorpublishedbyWileyPeriodicals,Inc. Correspondence SoledadAlcántara,DepartmentofPathology andExperimentalTherapeutics,Schoolof MedicineandHealthSciences,Universityof Barcelona,Barcelona,Spain. Email:salcantar[email protected] MartaCamprubíCamprubí,Neonatology Service,SantJoandeDéu,BCNatal,Hospital SantJoandeDéuiClínic,Universityof Barcelona,Barcelona,Spain. Email:[email protected] Present address JuanA.OrtegaCano,Department ofNeurology,FeinbergSchoolof Medicine,NorthwesternUniversity,Chicago, IL60611,USA Funding information Thisworkwassupportedbytwoprivate grantsonefromBebe′Due,SantaPerpetuade Moguda,Barcelona,Spainandtheotherfrom “Premi ambit de la infancia”AgrupacióMútua SpainandtoMCandbygrantsfromtheSpanish MinisteriodeEconomíayCompetitividad [MAT2011-29778-C02-02]cofinancedby theEuropeanRegionalDevelopmentFund, [BFU2014-55373-R]andfromtheGeneralitat deCatalunya[2014SGR1178]toSA.The sponsorhadnoinvolvementinthecollection, analysisorinterpretationofthedata Abstract Introduction: Intrauterine growth restriction (IUGR) is the failure of the fetus to achieveitsinherentgrowthpotential,andithasfrequentlybeenassociatedwithneurodevelopmentalproblemsinchildhood.Neurologicaldisordersaremostlyassociated withIUGRbabieswithanabnormallyhighcephalizationindex(CI)andabrainsparing effect.However,asimilarcorrelationhasneverbeendemonstratedinananimalmodel. TheaimofthisstudywastodeterminethecorrelationsbetweenCI,functionaldeficits inlearningandmemoryandalterationsinsynapticproteinsinaratmodelofIUGR. Methods: Utero-placentalinsufficiencywasinducedbymeso-ovarianvesselcauterization (CMO)inpregnantratsatembryonicday17(E17).Learningperformanceinanaquatic learningtestwasevaluated25daysafterbirthandduring10days.Somesynapticproteins wereanalyzed(PSD95,Synaptophysin)byWesternblotandimmunohistochemistry. Results: PlacentalinsufficiencyinCMOpupswasassociatedwithspatialmemorydeficits,whicharecorrelatedwithaCIabovethenormalrange.CMOpupspresentedalteredlevelsofsynapticproteinsPSD95andsynaptophysininthehippocampus. Conclusions: Theresultsofthisstudysuggestthatlearningdisabilitiesmaybeassociatedwithaltereddevelopmentofexcitatoryneurotransmissionandsynapticplasticity. Althoughinterspecificdifferencesinfetalresponsetoplacentalinsufficiencyshouldbe takenintoaccount,thetranslationofthesedatatohumanssuggestthatbothIUGR babiesandbabieswithanormalbirthweightbutwithintrauterineDoppleralterations andabnormalCIshouldbecloselyfollowedtodetectneurodevelopmentalalterations duringthepostnatalperiod. KEYWORDS cephalizationindex,intrauterinegrowthrestriction,learning,placentalinsufficiency,spatial memory,synapticplasticity 2 of 11 | CAMPRUBÍ CAMPRUBÍ et Al. 1 | INTRODUCTION Intrauterine growth restriction (IUGR) is the failure of the fetus to achieveitsinherentgrowthpotentialandcanbeduetomaternalor fetal problems (Goldenberg etal., 1989). The incidence of IUGR in developedcountriesisaround3–7%ofthetotalpopulation,butmay varydependingonthecriteriaused(Romoetal.,2009).Thedegrees anddefinitionsofIUGRdiffer,butrecentdiagnosticcriteriaconsider fetusestobetrulyIUGRwhentheestimatedweightislessthanthe 10th percentile for gestational age and on presenting an abnormal fetalDopplerultrasound.SevereIUGRisdefinedasweightbelowthe 3rd percentile (Figueras & Gardosi, 2011; Zhang, Merialdi, Platt, & Kramer,2010).However,accordingtothegrowthstandardsapplied, anumberofbabiesmaybeconsideredgrowthrestricted(notachievingtheirpotentialgeneticgrowth)butnotIUGR,becausetheirbirth weight(BW)iswithinthe10thpercentileoraboveandconsideredin thenormalrange. ChildrenbornwithIUGRarepredisposedtolong-termcognitive andneurodevelopmentalimpairmentthatisassociatedwithlearning difficulties and decreased academic performance (Arcangeli, Thilaganathan, Hooper, Khan, & Bhide, 2012; de Bie, Oostrom, & Delemarre-van de Waal, 2010). Longitudinal studies on IUGR babies have recently identified the cephalization index (CI=head circumference/BW)asoneofthebestperinatalparametersforthe prediction of neurodevelopmental outcome (Geva, Eshel, Leitner, Fattal-Valevski,&Harel,2006).AhighCIinIUGRbabiesseemsto beindicativeofintrauterinebloodflowredistribution,alsoknown asthebrain sparingeffect,andisstronglycorrelatedwithareductioninIQ,difficultiesincreativeproblemsolving,deficitsinattention andexecutivefunctions,andalterationsinvisuomotororganization (Flood etal., 2014; Van den Broek, Kok, Houtzager, & Scherjon, 2010). TheunderlyingmechanismsofneurologicimpairmentinIUGR remainunclear,althoughsomeauthorshavesuggestedthatthey maybe associatedwith structuralbrainalterations.Asignificant reductioninintracranialvolumeandcorticalgraymatterhavebeen reportedinIUGRbabies,andsomeauthorshavesuggestedthat a selective cortical vulnerability may be involved (Eixarch etal., 2012). Inthesameline,neuroimagingandhistopathologicalstudiesof theIUGRbrainininfantsandanimalmodelshaveshownthatone oftheareasmostaffectedisthehippocampus(Dieni&Rees,2003; Listeretal.,2005;Lodygenskyetal.,2008).Thehippocampusisa structureinvolvedincodification,storage,andrecoverymechanisms relatedtolearningandmemoryprocesses(Vann&Albasser,2011). Inthehippocampus,CornusAmmonis1,3(CA1,CA3)anddentate gyrusarethemostvulnerableregionstohypoxia-ischemia(HI),an inherentconditioninutero-placentalinsufficiency(Hsiao&Patterson, 2012).Themainstructuralchangesdescribedincludeareductionin cellnumberandwhitemattercontentandalteredcellularcompositionanddendritemorphology(Dieni&Rees,2003;Fungetal.,2012; Mallard,Loeliger,Copolov,&Rees,2000).Allthesechangeshavealso beenassociatedwithmodificationsinthegeneexpressionofneurotrophicfactors,glutamateneurotransmitterreceptors,andsynaptic proteinsthateventuallyleadtoimportantsynapticalterations(Chen etal.,2007;Dieni&Rees,2005;Schoberetal.,2009;Sommeretal., 2010). Theaimofthisstudywastoassesstheeffectofutero-placental insufficiencyonlearningandmemoryprocesses.WeusedaCMO animalmodelofIUGR(Camprubietal.,2009)todeterminethecorrelation between functional deficits in learning and memory and synapticproteinsexpressioninthehippocampus.Moreover,wealso studiedwhethertheCImightbeagoodmarkertoidentifypatients withabrain-sparingeffectdespitetheirBWbeingwithinthenormal range. 2 | MATERIALS AND METHODS 2.1 | CMO procedure and anthropometric analysis Thestudiesweredesignedtominimizethenumberofanimalsused in the procedure. All the animal protocols were approved by the Institutional Animal Care and Use Committee in accordance with SpanishandEuropeanUnionregulations. Thirty-twopregnantWistarrats(HarlamLaboratories)weremaintainedina12h-light/darkcyclewithad libitumaccessto standard rodentdietandwater.TheCMOprocedurewasperformedinpregnantratsatembryonicday17(E17)(n =17).Briefly,undergeneral anesthesia with isofluorane 2.5% dams underwent laparotomy and twomeso-ovarianvesselslocatedinthebordersbetweentheupper andmediumandthemediumandlowerthirdsofthecorresponding hornwerecauterized.Thisprocedurehasbeenvalidatedasaplacental insufficiencyinductionmethod(Camprubietal.,2009).Controlanimalsunderwentidenticalanesthesiaandbothuterinehornsandtheir vessels were exposed and examined with no further manipulation (n =15).TheCMOmodelproducesmoreuniformischemiathroughouttheuterusandthereforeaffectsalltheimplantedfetusesmore homogeneouslythantheuterinearteryligationmethoddescribedby Wigglesworth(1964).InCMOmodelBWwasnotdependentonthe positionofthefetusesandwasquite homogeneous,withasimilar proportionofpups<10thcentileineachlitter(Wigglesworth1964; Camprubíetal.,2009). The rats were allowed to deliver spontaneously and the pups werebornbetweendays21and22ofgestation.Toavoidrejection bythemother,newbornpupswerecountedandtheirBWrecorded ontheseconddayoflife(P2)(accuracy:0.1g;PB3001-LMettler Toledo). Theanimalswerethen divided intothreeexperimentalgroups with both sexes being similarly represented. Pups born to sham- operatedmotherswereconsideredascontrols(n =68).Accordingto humandiagnosticcriteria,pupsborntoCMO-operatedmotherswere divided into two different experimental groups: the animals most affected were denominated the IUGR group which was composed ofpupswithaBWbelowthe10thpercentile(<5.07g,n =31);and | 3 of 11 CAMPRUBÍ CAMPRUBÍ et Al. thosewithmildinvolvementbutwithanormalBW(≥10thpercentile) weredenominatedischemic(≥5.07g,n =42).Toensureahomogeneousmilksupplythroughoutthelitter,amaximumoffourratpups wereallowedpermother.Inthecaseofcontrols,onlythepupswitha normalBWwereincludedinthestudy.Aschemeofthenumberand proceduresperformedintheanimalsstudiedisshowninFigure1. Onday35(P35),theratswereweighedandthegrowthvelocity was calculated as: Toevaluatethebrain-sparing effect,ninecontrols,fourischemicand fiveIUGRpupsweresacrificedandanalyzedatP2and16controls, 15ischemicand14IUGRpupswerestudiedatP35.Thebrainswere rapidlyremovedandweighed.Takingintoaccountthedifficultyand lowaccuracyofheadcircumferencemeasuresinrats,theCI(calculatedinhumansasCI=headcircumference×100/BW(g))wascalculatedusingtheweightofthebrain.IntheseanimalsamodifiedversionofthehumanCIwascalculatedasCI=brainweight(mg)×100/ BW(mg).ThismodifiedversionoftheCIreflectsthebrain/BWratio, expressedasapercentage. 2.2 | Water maze test Spatialmemoryanalysiswascarriedoutaccordingtoaversionofthe Morris’ water maze (MWT) pool modified for small animals (63cm long,43cmwideand35cmhigh),withthewatertemperaturesetat 22–23°Candmadeopaquebylatexsuspension(Morris,1984).The escapelatency(EL),definedasthetimetakentoreachtheplatform, wasmeasuredduringeachtrialasanindicatoroflearning.Onpostnatalday25,thepupsofeachgroup(55controls,34ischemicand22 IUGR)weretrainedfor10consecutivedaysandthensacrificedatday 35.TheCIoftheseanimalswascalculatedanditsinfluenceonthe learningcurveanalyzed. 2.3 | Western blots (WB) AsshowninFigure1,agroupof50rats(14IUGR,15ischemicand 16 controls) were trained in the MWT from P25 to P35 (trained animals), four animals from each group (n =12) were not trained (untrainedanimals).Theratsweresacrificedatday35bydecapitationandtheirbrainswererapidlyremoved,weighedandprocessed forWBanalysis. Protein extracts from the hippocampus were separated by SDS-PAGE and electro-transferred to a nitrocellulose membrane. Membranes were blocked with 5% nonfat dry milk in Tris-buffered saline and incubated first with primary antibodies overnight at 4°C (1:2000, PSD95; Abcam, Cambridge, UK; 1:1000; Synaptophysin, Dako, Glostrup, Denmark; 1:10,000, goat anti-Actin, 1:10,000; Santa CruzBiotechnology,SanDiego,CA,USA),andthenwiththeircorresponding secondary HRP-conjugated antibodies (1:5000, Santa Cruz Biotechnology).TheproteinsignalwasdetectedusingtheECLchemiluminescentsystem(Amersham,Buckinghamshire,UK).Quantificationof thelabelingintensityinmicrographsanddensitometryanalysisforWB bandswascarriedoutusingImageJsoftware(Rasband,W.S.,ImageJ, U.S.NationalInstitutesofHealth,Bethesda,MD,USA,http://rsb.info. nih.gov/ij/,1997–2009).Histogramandgeltoolswereusedtoanalyze confocalmicrographsandWBbands,respectively.Alltrained(14IUGR, 15ischemicand16controls)anduntrained(4pergroup)animalswere analyzed,loading3–4animalspergroupandgel,followedbyacomparativestudy. 2.4 | Histological and immunohistochemistry analysis Forhistologicalandimmunohistochemistryanalysis,trainedrats (8IUGR,19ischemicand39controls)werekilledbyanesthetic FIGURE1 Graphicalschemeoftheexperimentalgroupsandprocedures.WMT:watermazetest [Weight at P35 (g) −weight at P2 (g)]∕33(days). 4 of 11 | CAMPRUBÍ CAMPRUBÍ et Al. overdoseandtranscardiallyperfusedwith4%paraformaldehyde (PFA).Dissectedbrainswerepostfixedfor6hin4%PFA,cryoprotectedandkeptfrozen.Coronalsectionsof40-μmthickwerecollectedinacryoprotectivesolution(40%0.1Mphosphatebuffer, 30%glycerol,and30%ethyleneglycol),andstoredat−30°Cfor furtheruse. Forbroadhistologicalanalysistissuesectionsweremountedon polylysinecoatedslicesandstainedwithHematoxylin–Eosin(H-E). Immunohistochemistrywasperformedintissuesectionsblocked with10%methanolinPBSfor15minfollowedby0.4tritonX-100 PBS+10% goat/horse normal serum (Gibco-Life Technologies refs. 16210–072/16050–122) for 2hr and subsequently incubated with primary antibodies at 4°C overnight. The sections werethenincubatedwithbiotin-conjugatedsecondaryantibodies (1:200,Vector, Burlingame, CA, USA) for 1hr and subsequently with a Streptavidin-peroxidase complex (1:400; Amersham) for 2hratroomtemperature.Theenzymaticreactionwasdeveloped with diaminobenzidine (DAB, ref. D123384-SG; Sigma-Aldrich Co,SantLouis,MO,USA)andthesectionsweredehydratedand coverslippedwithEukitt(Panreac,Barcelona,Spain).Primaryantibodiesagainst thefollowingproteinswereused:GFAP(1:1000, ref.Z0334Dako),NeuN(1:500,ref.MAB377,Millipore,Tamecula, CA, USA) PSD95 (1:1000, ref ab18258,Abcam), Synaptophysin (1:1000,refIR776,Dako). MicrographswerecapturedwithalightmicroscopeNikonEclipse 800(Nikon,Tokio,Japan)andassembledinAdobePhotoshop(v.7.0), withthesameadjustmentsforcontrast,brightness,andcolorbalance toobtainoptimumvisualreproductionofthedata.ThenumberofCA1 pyramidalneuronswascountedatthelevelofthedorsalhippocampus in400×micrographsofNeuN-stainedsectionswiththeaidofImageJ software.Fourserialpicturestaken300micronsapartofthreeanimals pergroupwereused. 2.5 | Data analysis ThedailyELwasaveragedfromfourtrainingtrialsandanalyzedusing multifactorANOVA(time,groupandsex)andtheFisher’sleastsignificantdifference(LSD)test.ANCOVAtestwasusedtoanalyzetheinfluenceoftheCIandBirthweightonthelearningprocess.Differences incellnumber,CI,growthvelocity,andsynapticproteinexpression amongthethreegroupswerecalculatedusingone-wayANOVAand theFisher’sleastsignificantdifference(LSD)test.ThestatisticalanalysiswasperformedusingStatgraphicsCenturionprogram. 3 | RESULTS 3.1 | Birth weight and growth rates of the CMO and control groups According to experimental manipulation and BW, the pups were dividedintoonecontrolgroupandtwoCMOgroups:ischemicand IUGR; depending on their birth weight. Males and females were equallyrepresentedineachgroup(p=.73). Someoftheanimalswerekilledatday2(seeTable1foranthropometricdata). TheremaininganimalswerekilledatP35andsomewerestudiedby WB,withthebrainsoftheseanimalsbeingextractedinfreshandweighed (seeTable2foranthropometricdata).Theotheranimals,usedforhistology,wereperfusedandtheirbrainswerenotweighed.Asreportedpreviously(Camprubietal.,2009),theCIofthenewbornIUGRpupswas significantlyhigherthanthatofthecontrolanimals(p<.01).Surprisingly, althoughtheBWoftheischemicpupswaswithinthenormalrange(≥10th percentile),theirCIwassignificantlyhigherthanthatofthecontrols,but lowerthantheIUGRgroup(p<.01,Table1).Anthropometricparameters werealsorecordedinanimalsatP35inordertoevaluategrowthvelocityandtheCI(Table2).Inlinewithpreviousstudiesbyourgroup,byP2 nodifferencesinBWwereobservedbetweensexes.However,atP35 IUGRfemalesweresignificantlysmallerthanIUGRmales(p<.05).The growthvelocityinbothCMOgroupswassimilarandsignificantlyslower thaninthecontrolgroup(p<.01,Table2),andthus,thetwoCMOgroups weresmaller(21–26%respectively)thanthecontrolsatP35.Duetothe allometricgrowthoftheheadandbody,theCIatP35wasconsiderably smallerthanatP2,andneonataldifferencesamonggroupsweremaintainedalongthepostnataldevelopmentwithischemicandIUGRgroups presentingasignificantlyhigherCIthanthecontrols(p<.01)(Table2). ACIover1.32was observedin93% and 87% ofischemic and IUGR pups,respectively,butonlyin25%ofthecontrolpups.Onanalyzingthe correlationbetweenBWandtheCIatP2,therhovaluewas−0.9599, suggestinganinversecorrelationbetweenBWandtheCI.Thisrelation wasalsoinverseatP35,being−0.7144.Thereisnostatisticalsignificant correlationbetweenFinalELandCIrho:−0.2049. 3.2 | Effect of uteroplacental insufficiency in spatial memory performance TheMWTmodifiedforyounganimalswasusedtoevaluatelearning andspatialmemory.TheELofeachanimalwasrecordedfourtimes Control N = 9 Ischemic N = 4 IUGR N = 5 Birthweight(P2)(g) 6.13±0.39 5.66 ± 0.27*4.45 ± 0.13**,## Brainweight(P2)(g) 0.24 ± 0.02 0.27 ± 0.01** 0.26 ± 0.01 Cephalizationindexatday2 3.99±0.32 4.86 ± 0.23** 5.87 ± 0.07**,## IUGR,Intrauterinegrowthrestriction.Inthetable,Birthweight,brainweightandcephalitzationindex atdaytwoispresented. Eachgroupbelongstoatleasttwodifferentmothers.Dataareindicatedasmean±SD. Significantdifferencescomparedtocontrols**p<.01,*p<.05;toIschemicrats##p<.01. TABLE1 Theanthropometricparametersofnewbornrats | 5 of 11 CAMPRUBÍ CAMPRUBÍ et Al. aday,for10consecutivedays.Theresultswere analyzedbymultivariateANOVA withrespectto trialday,experimental groupand gender,withallthreeparameterssignificantlyaffectingthelearning process(p=.00001,p=.0008andp=.0016,respectively).Theperformanceofthecontrols,andtheischemicandIUGRgroupsimproved overtime(p<.01),butsignificantdifferenceswereobservedintheir learningprogression(Figure2a).Alongthelearningprocess,control ratswerebetter(shorterEL)thanischemicandIUGRrats(p<.01and p<.05,respectively).Nosignificantdifferenceswerefoundbetween theischemicandIUGRgroups(p=.672).FinaltheELvaluesofthe ischemicandIUGRanimalsremainedsignificantlyhigherthanthose ofthecontrols(controls:10.2±6.3s;ischemic:16.7±12.9sp<.05; IUGR:19.3±12.3s,p<.01;mean±SD). We considered gender as a relevant factor in this analysis. However,nosignificantdifferenceswerefoundinthelearningprocessbetweentheischemicandIUGRgroups.Therefore,weclustered thetwoCMOgroupstogetherandrepeatedtheanalysissegregating thecontrolsversustheCMOgroupsbysex(Figure2b).Thelearning curves showed significant differences according to gender and the CMOprocedure(p=.002andp=.0034,respectively).Despiteintergroupdifferencesinthelearningcurves,themalesandfemalesinthe differentgroupsachievedsimilarfinalELscores(control:♂9.9±5.9s; ♀ 10.7±6.6s; CMO: ♂ 18.0±10.9s; ♀ 17.5±13.2 s; mean±SD). TheCMOprocedureaffectedbothgenderssimilarly(p=.6290),and the scores were significantly poorer than their respective controls (♂p=.01and♀ p=.05LSDtest). Anthropometric parameters (Table2), including body and brain weight,growthvelocityandtheCIwereanalyzedaspossiblecovariablesinthelearningprocess.ThelearningprocessshowedasignificantassociationwiththeCI(males,p=.0039;females,p=.0240)but notwithBW(males,p=.671;females,p=.469).Surprisingly,there wasnocorrelationbetweentheCIandMWTperformance(p=.213). Independentlyoftheexperimentalgroup,CIvaluesabove1.32were associatedwithanincreasedriskofpoorspatialmemoryperformance (ELbelowthe25thpercentile,estimatedinthecontrolgroup),ranging from33%incontrols(1of3)to58%inpupsfromCMO-operated mothers(15of26)(Figure2c). These results suggest that a high CI associated with utero- placentalinsufficiencypredisposedthepupsstudiedtolearningand memorydeficits. 3.3 | Effect of uteroplacental insufficiency on cell number and synaptic protein expression in the hippocampus At P35, the hippocampus of ischemic and IUGR rats were normal without gross histopathological differences in H-E (Figure3a–c) NeuN(Figure3d–f),andwithmildgliosisobservedinGFAPstaining (Figure3g–i).However,asreductionsinthenumberofhippocampal cellshavebeendescribedinotherIUGRanimalmodels(Mallardetal., 2000), we performed an in depth analysis of the impact of CMO- induced utero-placental insufficiency on the number of neurons immunolabeledwithNeuNinthepyramidallayeroftheCA1subfield ofthe hippocampus(n =3from eachexperimentalgroup). Asmall, albeitstatisticallysignificant,differencewasfoundinthetotalnumberofneuronsintheCA1pyramidallayerintheischemicandIUGR groups compared to controls (control=65±9; ischemic=54±7, p<.001;IUGR=48±4;mean±SD,p<.001). Toevaluatetheintegrityofthefunctionalsynapticmachineryin thehippocampus,totalproteincontent,andthedistributionoftwo specificsynapticproteins,postsynapticdensityprotein95(PSD95) andsynaptophysin,wereanalyzedbyWBandimmunofluorescence imaging(Figure4a,b).PSD95isascaffoldingproteininvolvedinglutamate receptor clustering and postsynaptic intracellular signaling TABLE2 Evolutionoftheanthropometricparametersintheexperimentalgroups Control CMO Control (N = 68) Ischemic (BW > 10%, N = 42) IUGR (BW < 10%, N = 31) Male (N = 37) Female (N = 31) Male (N = 20) Female (N = 22) Male (N = 12) Female (N = 19) Weight(g)atday2 6.35±0.79 [5.07–7.56] 5.91±0.52 [5.30–6.83] 6.3 ± 0.43 [5.12–6.51] 5.83±0.44 [5.58–6.73] 4.39 ± 0.50**,## [4.12–4.87] 3.32 ± 0.648**,## [3.43–4.9] &Male (N = 7) Female (N = 9) Male (N = 6) Female (N = 9) Male (N = 8) Female (N = 6) Weight(g)atday35 144.09±17.5 130.2±9.3 118.10 ± 14.22** 105.70 ± 15.99** 108.44 ± 14.50** 91.02 ± 16.14§,** Brainweight(g)at day35 1.65 ± 0.12 1.60±0.08 1.70±0.08 1.60 ± 0.07§1.54 ± 0.10*,## 1.47 ± 0.04**,## Growthvelocity (g/day) 4.17 ± 0.50 3.75 ± 0.27 3.40 ± 0.42** 3.02 ± 0.50** 3.15 ± 0.44** 2.63 ± 0.49**,§ Cephalizationindex atday35 1.16 ± 0.12 1.23 ± 0.10 1.46 ± 0.15** 1.55 ± 0.24*1.44 ± 0.12** 1.65 ± 0.29§,** IUGR,Intrauterinegrowthrestriction;BW,Birthweight.Inthetableevolutionofsomeantthropometricalparametersispresented. Eachgroupbelongstoatleasttothreedifferentlitters.Dataareindicatedasmean±SD. &:Theseparametersareobtainedfromthoseanimalsusedforproteinextraction. Significantdifferences(LSDtest)comparedtocontrol**p<.01,*p<.05;toischemic#p<.05;tomale§p<.05. 6 of 11 | CAMPRUBÍ CAMPRUBÍ et Al. (Elias& Nicoll,2007), andsynaptophysinis anintegralmembrane vesicularsynapticglycoproteinimplicatedinvesiclebiogenesisand recycling (Valtorta, Pennuto, Bonanomi, & Benfenati, 2004). WB analysesrevealedthatinbothtrainedanduntrainedanimalsthetotal amountofPSD95proteinwassignificantlyreducedinthehippocampusofIUGRanimalscomparedtothecontrolandischemicgroups. Incontrast,PSD95proteincontenttendedtoincrease,albeitnotsignificantly,inischemicanimals.Asignificantreductionwasalsofound inthe totalamountof synaptophysin proteinin the hippocampus of untrained IUGR animals compared to the control and ischemic groups. However, the total amount of synaptophysin proteinwas normalizedinthehippocampusofIUGRratsafterwatermazetraining(Figure4a).Immunohistochemistryanalysisofthehippocampus oftrainedratsrevealedasimilarlaminardistributionofPSD95and synaptophysinstaininginthethreegroups(Figure4b).However,in linewiththereductioninPSD95proteinobservedinWB,theintensityofPSD95immunostainingwasreducedinthehippocampusof theIUGRgroup. FIGURE2 Learningprogressionandcorrelationbetweenescapelatencyandthecephalizationindex(CI).(a)Plotrepresentingthe averageescapelatencyinfourtrialsperformedeachdaybythecontrolsandtheischemicandIntrauterinegrowthrestriction(IUGR)groups (n =55,n =34,N =22,respectively,withsimilarrepresentationofbothgenders)duringthe10daysofthestudy.Learningprogression fittologarithmiccurvesandANOVAanalysisrevealedsignificantdifferencesintheshapeofthecurveandthefinalscoresbetweenthe controlsandtheIUGRandischemicrats(p<.001),withnosignificantdifferencesbetweenIUGRandischemicrats.(b)Plotrepresentingthe averageescapelatencyforthesameanimalsasin(a),butsegregatedbygenderanddistributedintocontrolsandCMO(IUGR+ischemic) experimentalgroups.MultifactorANOVAandLSDanalysisrevealedsignificantdifferencesintheshapeofthelearningcurvesbetween malesandfemales(p<.01LSDtest),andbetweentheircorrespondingcontrolsandCMOgroups(malesp<.05,femalesp<.01,LSDtest). (c)Dispersionplotsrepresenttheaverageescapelatencyofeachindividual(controlmale=7,controlfemale=9;ischemicmale=6;ischemic female=9;IUGRmale=8;IUGRfemale=6)withrespecttoitsCIonthreerepresentativedaysofthestudy.Verticalandhorizontallineson day10indicatethe75thpercentileofthecontrols(CI=1.32;escapelatency[EL]=17.7s).Errorbarsinaandbrepresentthestandarderror ofthemean | 7 of 11 CAMPRUBÍ CAMPRUBÍ et Al. 4 | DISCUSSION The main findings of this study suggest that neurodevelopmental alterationsaffectmalesandfemalessimilarlyandarenotonlyassociatedwithBWbutalsowithanabnormallyhighCIandalteredsynaptic machineryinthehippocampus. IUGRisoneofthemostfrequentandcomplexconditionsinmodernobstetrics,beingmainlyproducedbyplacentalinsufficiencyand undernourishment(Goldenbergetal.,1989;Maulik,2006).TheincidenceofIUGRhasbeenontheriseworldwideinthelastdecades (Romoetal.,2009).Themostfrequentlystudiedlong-termrepercussionsofIUGRarethoseinvolvingmetabolicprogramming,whichlead totheso-calledmetabolicsyndrome that is associatedwith a high riskofdevelopinghypertension,diabetes,andcardiovasculardisease. Nonetheless,longitudinalstudieshavealsorevealedahighincidence oflearning,schoolandbehavioralproblemsinIUGRbabiescompared withbabiesbornatthesamegestationalagebutwithanadequate BW(Leitneretal.,2007;Nardozzaetal.,2012).Itisimportanttonote thatthediagnosisoftrueIUGRremainsachallengeinclinicalpractice sincechildren<10thpercentileinsizearenotalwayspathological,and somechildrenwithappropriateweightfortheirgestationalagemight nothaveachievedtheirtargetgrowth(Campbelletal.,2012;Zhang etal.,2010).Inparticular,theunderlyingneuropathologyofcognitive andbehavioraldeficitsassociatedwithIUGRislargelyunknown,and theidentificationofchildrenatriskofpoorneurologicaloutcomeis verydifficultatearlyages(Leitneretal.,2007;Streimishetal.,2012). IUGR fetuses develop some hemodynamic adaptations during fetal life,redistributingblood flowtovitalorganssuch as theheartand brain.Inthesebabies,thesizeoftheheadisrelativelysparedwith respecttothegrowthofthebody,resultinginanabnormallyhighCI. Althoughthisbrainsparingeffectwasinitiallyconsideredtobeaprotectivemechanism,itisnowlargelyconsideredtobeanearlyindicatorofbraindistress(Hernandez-Andrade,Serralde,&Cruz-Martinez, 2012;Yanney&Marlow,2004)associatedwithanincreasedriskof alteredneurologicaloutcome(Figuerasetal.,2011;Gevaetal.,2006). FIGURE3 Hippocampushistology.Coronalhippocampalsectionsofcontrol(a,d,g,)ischemic(b,e,h)andIntrauterinegrowthrestriction (IUGR)(c,f,i)ratsatP35.(a–c)Nisslstained;d–f)immunostainedagainstNeuN,insetcorrespondtoahighermagnificationoftheCA1pyramidal layer;(g–i)GFAPimmunostained.CA1,CA2,CA3hippocampalfields;DGdentategyrus;H,hilus;SO,stratumoriens;SP,stratumpyramidale; SR,stratumradiatum;SG,stratumgranulosum;SLM,stratumlacunosum-moleculare;SM,stratummoleculare;Scalebar:150μm (a) (b) (c) (f) (e) (d) (g) (h) (i) 8 of 11 | CAMPRUBÍ CAMPRUBÍ et Al. Consequently,theCIhasbeenproposedasaclinicalparameterfor theearlydetectionofbabiesatriskofabnormalneurodevelopment (Fattal-Valevskietal.,2009;Rozaetal.,2008). Tounderstandthecorrelationbetweenlearningdeficitsandthe morphological,cellularandmolecularalterationsdevelopedinIUGR patients,manyanimalmodelshavebeendeveloped.Rodents(mice, rats,andguineapigs)arethemostcommonlyusedanimalmodelsto asses alterations associated with IUGR due to the high number of functional tests available that allow the evaluation of neurological outcome. Inthepresentwork,weusedamodelofchronicutero-placental insufficiencyinducedbytheCMOprocedureatE17.Thismodel produces relatively uniform ischemia throughout the uterus and resultsina42%survivalrateandIUGRwithasignificantbrainsparingeffectinthepups(Camprubietal.,2009).Asinothervascular models performed at day 17 of gestation (Huizinga etal., 2004) andatE18(Jansson&Lambert,1999;Ogata,Bussey,LaBarbera, &Finley, 1985), animals born to CMO mothers did not achieve catch up growth. Simmons, however, reported catch up growth inanimalsobtainedfromoperatedmothersatdayE19(Simmons etal.,2001). Toourknowledge,thismodifiedversionoftheCIusingthebrain weighthasnotbeenusedinotheranimalmodels.However,inarabbit modelcreatedbyEixarchandcolleges,theCIwasalsohigherinIUGR, being3.19versus2.5incontrolanimalsandsuggestingthattheintrauterineprocessofgrowthrestrictionissimilar(Eixarchetal.,2009). Toanalyzedifferencesinlearningandmemoryprocessesbetween severely and mildly growth restricted rats, pups born to CMO- operatedmothersweredividedaccordingtheirBWintoIUGR(<the 10thpercentile),andischemic,BW(>the10thpercentile)andwithin thenormallimits.DespitethedifferentBWs,thetwoCMOgroups showedasignificantlyhigherCIatbirthandalowergrowthvelocity comparedtocontrolstherebyreflectingtheabsenceofcatchup.The brainsparingeffectobservedinCMOpupsatbirth(higherCI)was maintainedthroughoutpostnataldevelopment,atleastuntilday35. Learning evolution was analyzed by the MWT, a specific hippocampal-relatedspatialmemorytest,andwasfoundtobesimilar intheCMOgroups.Thistesthastheadvantagesofrapidacquisition; performanceislessinfluencedbyBWanddoesnotrequirefoodor waterdeprivationasinothertestsandcouldbeadrawbackinrapidly growingpostweaningrats(Barrow&Leconte,1996;Ehman&Moser, 2006).The earlypostweaningperiodinrats(P25-P35)waschosen FIGURE4 Distributionofsynapticproteinsinthehippocampus.(a)WesternblotofPSD95,synaptophysinandSNAP25proteinsin hippocampaltissuefromcontrols,andtheischemicandIntrauterinegrowthrestriction(IUGR)groupsofuntrainedandwatermazetrainedrats. Actinwasusedastheproteinloadingcontrol.PSD95andsynaptophysinexpressionweresignificantlyreducedintheuntrainedIUGRgroup, whereasonlyPSD95expressionwasreducedaftertraining.(b)Photomicrographsofcoronalsectionsofthehippocampusoftrainedcontrol, ischemicandIUGRratsimmunostainingofpostsynapticproteinPSD95,andpresynapticproteinsynaptophysin.CA1,CA3,hippocampal subfields;DG,dentategyrus;H,hilus;SO,stratumoriens;SP,stratumpyramidale;SR,stratumradiatum;SG,stratumgranulosum;SLM,stratum lacunosum-moleculare;SM,stratummoleculare;*p<.05respectcontrol,##p<.01respectischemic;LSDtest.Scalebar:185μm | 9 of 11 CAMPRUBÍ CAMPRUBÍ et Al. becausethisperiodrepresentsamilestoneintheprogressiontoadult functioninmammalsandconstitutesanessentialstepinlearningperformanceinmanytests(Ehman&Moser,2006). Ingeneral,learningcurveswereslightlydifferentbetweenmales and females, although both genders achieved similar final scores accordingtotheirexperimentalgroup.Althoughsexdifferenceshave been suggested in the outcomes of animals subjected to neonatal hypoxia-ischemia (Smith, Alexander, Rosenkrantz, Sadek, & Fitch, 2014),thisisnotsupportedbyourdataandwatermazeperformance wasaffectedsimilarlybytheCMOprocedureinbothsexes.IUGRand ischemicanimalsofbothsexesneededmoretimetolearnthesame taskthantheirrespectivecontrols,andalthoughtheirperformance improvedovertime,theirfinalscoresremainedslightlylower. OuranalysisalsoshowedacorrelationbetweenCI>1.32(correspondingwiththe75thpercentileofthecontrolgroup)andpoorspatialmemory(belowthe25thpercentileofthecontrolgroup),although abovethisthreshold,thelargestCIsatP35werenotnecessarilyassociatedwiththepoorestscores.Onepotentialexplanationforthislack oflinearcorrelationbetweenCIandspatialmemoryperformancemay bethat,withincreasingseverityofutero-placentaldysfunctionand nutrient deprivation, critical adjustments in fetal brain metabolism might impair postnatal growth and neurodevelopmental outcome, masking the initial brain sparing effect and the increase in the CI. Althoughfurtherstudiesandasystematicanthropometricanalysisof individualsareneededtoaddressthisquestion,arecentsystematic reviewofanimalmodelsthatuseduterinearteryligationtoinduce IUGR reached similar conclusions. Fetuses with growth restriction demonstrateareductioninBWwithrelativesparingofbrainmassand deficitsinmanyneuroconductualtests(Basilious,Yager,&Fehlings, 2015). Moreover,thefunctionalalterationsdescribedinCMOanimalsin thisstudyareinlinewiththosedescribedbyLeitnerandcolleagues inacohortofIUGRbabies.TheseauthorsreportedthatIUGRbabies presentalterationsinspatialorientationandspatialmemory(Leitner, Heldman,Harel,&Pick,2005).Inaddition,typicalhippocampaldeficits,suchasexecutiveshort-termmemorydeficitscharacteristicof anterior hippocampal-prefrontal network alteration have also been reportedinIUGRbabies(Gevaetal.,2006).Thesimilarpoorspatial memory performance shown by ischemic and IUGR rats might be relatedtotheirparallelanthropometricevolution,asdescribedina cohortofIUGRpatients,inwhomthepresenceofcatchupgrowth wasrelatedtobetterneurologicalprognosisandanincreasedCIwith learningandmemorydisabilities(Fattal-Valevskietal.,2009;Leitner etal.,2007).Ourdatasuggestthatincasesinwhichutero-placental insufficiencyissuspected,ahigherCI(estimatedtobeabovethe75th percentile)maintainedduringchildhoodcouldbeabetterindicatorof suboptimalintrauterinegrowththanonlylowBW.Moreover,ahigh CIvaluecouldalsobeusefulasamarkerofincreasedriskofcognitive andneurobehavioraldeficits. Spatial memory plays an important role in the consolidation of learning and is a major part ofwhat is known as explicit memory or conscious memory of events that have occurred in the external world.Inmammals,thisprocesstakesplaceintheCA1regionofthe hippocampus, and the underlying cellular mechanism is a form of synapticplasticityknownaslong-termpotentiationorLTP(Vann& Albasser,2011),thatrequiresCa+2signaling,geneexpressionchanges andproteinsynthesis(Morgado-Bernal,2011).Theneuropathological basesoflearningandmemoryalterationsinIUGRbabiesarenotclear, butthereisconsensusthatwithinthebrainthehippocampusisoneof theareasmostsusceptibletoIUGRandhypoxicdamage(Fungetal., 2012;Lodygenskyetal.,2008;Mallardetal.,2000),whichparticularly affectsCA1pyramidalneuronsinthehippocampus(Kovalenkoetal., 2006;Listeretal.,2005).Inthisstudy,theischemicandIUGRCMO groupspresentedareductioninCA1neuronalnumber,mildgliosis and paradoxical changes in PSD95 and synaptophysin expression. PSD95isaneuronalproteinanchoredinthepostsynapticcompartmentoftheglutamatergicneurons,whereitactsasascaffoldingproteinnecessaryforAMPARandNMDARglutamatereceptortrafficking and transsynaptic signaling, thereby modulating excitatory synaptic strength,excitatory/inhibitorybalance,andplasticity(Elias&Nicoll, 2007;Keith&El-Husseini,2008).AbnormallyreducedPSD95expressionimpairsactivity-dependentsynapticstabilization(Ehrlich,Klein, Rumpel,&Malinow,2007),whereasPSD95overexpressionincreases theexcitatorytoinhibitorysynapsesratioandaltersdendriticarbor development(Bustosetal.,2014;Prange,Wong,Gerrow,Wang,&El- Husseini,2004).PSD95wasdramaticallyreducedinthehippocampus ofIUGRrats,similartowhathasbeendescribedinanimalsexposed toperinatalhypoxiawhichalsodisplayedimpairedhippocampalmaturationanddeficitsinspatiallearningandmemory(Chenetal.,2006, 2007). Abnormal PSD95 expression has also been associated with autismspectrumdisordersinhumansandanimalmodels(Keith&El- Husseini,2008;Tsaietal.,2012),andthesedisordershavealsobeen relatedtoIUGR(Leonardetal.,2008;Walkeretal.,2015).Presynaptic machinery is also affected by utero-placental insufficiency and is shownbyreducedproteinlevelsofsynaptophysininthehippocampus ofnaiveIUGRratsandtherecoveryofnormallevelsaftertraining. Synaptophysin is a presynaptic protein involved in synapticvesicle biogenesisandtraffickingthatincreasesinparallelwiththeformation ofsynapsesandisfrequentlyusedasanindicatoroftotalsynapticvesiclespools(Valtortaetal.,2004).Itisknownthatpositivestimulisuch asphysicalexerciseandenvironmentalenrichmentincreasesynaptic proteinsexpressionandnormalizeinterneurondevelopment,andare associatedwithimprovedsynapticplasticity and cognitivefunction (Abel&Rissman,2013;Chen,Chen,Lei,&Wang,1998;Chenetal., 2006; Hu,Ying, Gomez-Pinilla, & Frautschy, 2009; Komitova etal., 2013).Therefore,itispossiblethatexerciseinherenttowatermaze trainingmightaccountforthesynaptophysinnormalizationinIUGR ratsaftertrainingsuchaswhatwasobservedinthisstudy,andsuggeststhatenvironmentalmanagementmightreversesomeofthesynapticchangesassociatedwithprenatalgrowthrestrictionpostnatally. 5 | CONCLUSIONS In this study, a murine model was used to determine part of the biologicalmechanismsoflearningandmemorydisordersrelatedto