neural encoding of olfactory recognition memory › 6d14 › b1d4cb3ab7dd... · neural encoding of...

Journal of Reproduction and Development, Vol. 51, No. 5, 2005

—Review—

Neural Encoding of Olfactory Recognition Memory

Gabriela SÁNCHEZ-ANDRADE1), Bronwen M JAMES1) and Keith M KENDRICK1)

1)Cognitive and Behavioural Neuroscience, The Babraham Institute, Babraham, Cambridge CB2 4AT, UK

Abstract. Our work with both sheep and mouse models has revealed many of the neural substratesand signalling pathways involved in olfactory recognition memory in the main olfactory system. Adistributed neural system is required for initial memory formation and its short-term retention–theolfactory bulb, piriform and entorhinal cortices and hippocampus. Following memory consolidation,after 8 h or so, only the olfactory bulb and piriform cortex appear to be important for effective recall.Similarly, whereas the glutamate–NMDA/AMPA receptor–nitric oxide (NO)–cyclic GMP signallingpathway is important for memory formation it is not involved in recall post-consolidation. Here,within the olfactory bulb, up-regulation of class 1 metabotropic glutamate receptors appears tomaintain the enhanced sensitivity at the mitral to granule cell synapses required for effective memoryrecall. Recently we have investigated whether fluctuating sex hormone levels during the oestrouscycle modulate olfactory recognition memory and the different neural substrates and signallingpathways involved. These studies have used two robust models of social olfactory memory in themouse which either involve social or non social odours (habituation-dishabituation and socialtransmission of food preference tasks). In both cases significant improvement of learning retentionoccurs when original learning takes place during the proestrus phase of the ovarian cycle. This isprobably the result of oestrogen changes at this time since transgenic mice lacking functionalexpression of oestrogen receptors (ERα and ERβ, the two main oestrogen receptor sub-types) haveshown problems in social recognition. Therefore, oestrogen appears to act at the level of the olfactorybulb by modulating both noradrenaline and the glutamate/NO signalling pathway.Key words: Olfactory bulb, Olfactory memory, Olfactory recognition, Social learning, Nitric oxide(NO), Oestrous cycle, Sheep, Mouse, Social transmission of food preference

(J. Reprod. Dev. 51: 547–558, 2005)

or several decades, olfactory learning has beenused as a model for the study of memory. This

is not only because mammalian olfactory learningshows a number of features common to the humandeclarative memory system one—trial learning,rapid encoding, long lasting memories and a highcapacity for information storage–but also becauseolfactory memory models represent a relativelysimple form of learning and the olfactory system

has well characterised pathways and structuralorganisation which offers many advantages indefining how processing occurs at a systems level.This makes it feasible to study plasticity at all levelsin the system, from the initial sensory detection ofodourant molecules at the level of the olfactoryreceptors r ight through to higher cort icalprocessing.

There are many different forms of olfactorylearning and memory in diverse behaviouralcontexts involving different mechanisms within theolfactory system. Olfactory learning can be divided

Accepted for publication: June 2, 2005Correspondence: K.M. Kendrick(e-mail: [email protected])

SÁNCHEZ-ANDRADE et al.548

into socially and non-socially motivated tasks–adistinction that has been shown to be important bywork on the oxytocin knockout mouse [1] whereonly deficits in social memory have been observed.The importance of olfaction and olfactory memoryo n s o c i a l l y m e d i a t e d b e h a v i o u r s i s v e r yconsiderable since many mammalian species relyupon smell to obtain information about theirenvironment, particularly their social one. Ourwork for the past 15 years has focused on theinvolvement of the main olfactory system in socialolfactory learning in the context of lamb-motherewe bonding via odour recognition [2, 3] and socialrecognition of individuals [4, 5] and socialtransmission of food preference [6, 7] in rodents.

These paradigms represent good ethologicallyrelevant models for the study of social olfactorymemory. Offspring recognition ensures that themother will preserve her own genes by limitingmaternal investment to her own offspring. Socialrecognit ion of conspecif ics has funct ionalconsequences in several social contexts [3, 8–11]. Itis crucial for the determination and maintenance ofsoc ia l s t ruc tu re in m a ny a n ima l spe c i e s ,particularly those living in small, stable socialgroups. With social transmission of foodpreference, after smelling a novel food odour on aforager’s breath rodents seek out that foodassuming that it is good to eat and suppressingtheir normal neophobic response. These modelsalso offer many advantages in studying olfactorymemory processes since learning is rapid androbust and long-lasting memories can be formedthat are relatively simple to test experimentally.

In sheep, the maternal ewe is able to recognise itsoffspring by identifying the individual body odourof her own lamb(s). She forms a selective olfactorymemory for each of her offspring within 2–4 h ofparturition and will consequently reject theapproach of any strange lamb [12]. Both thesensitive period for odour learning and thematernal acceptance behaviour are dependent onthe hormonal environment and are triggered byfeedback to the brain from mechanical stimulationof the vagina and the cervix which normally occursduring parturition. Thus, a ewe can be induced toaccept, and form, a new recognition memory for astrange lamb by simply mimicking birth throughadminis ter ing a few minutes o f ar t i f i c ia lvaginocervical stimulation (VCS) for up to threedays post-partum [13].

Social recognition tests in rodents rely on theintrinsic motivation they have for investigatingnovel conspecifics when introduced into familiarterritory or home cage. Under these circumstancesthey will investigate a novel conspecific more thana familiar one since they are already acquaintedwith the odour cues of the latter rather than theformer [14]. Odour cues such as soiled bedding orurine from familiar animals are sufficient for therecognition process [15]. In the social recognitiontest [16], a rat or mouse is exposed to a novelstimulus animal (generally juveniles) on twooccasions, separated by a given time interval. Thereduction of investigation time during the secondencounter is taken as evidence for social memory.Introduction of a new fresh stimulus animaltriggers an intense investigation bout similar to theinitial encounter, proving that the reducedinvestigation is not due to generalised habituation.This paradigm has been used mainly as a model ofshort term memory since it was found to last nolonger than 1 h in studies using individuallyhoused rats or mice [16–18]. However, recentstudies have shown that housing conditions affectretention of these social memories since micehoused in groups, rather than individually, wereshown to remember a familiar juvenile for up to 7days [19].

We have combined two versions of the originalsocial recognition paradigm in mice. In the firstinstance a habituation-dishabituation paradigm hasbeen employed [20, 21] where the test animal isexposed repeatedly (four trials) to a stimulusanimal (anaesthetised adult) for 1 min with 10 mininter-trial intervals. This results in a progressivegeneral reduction of investigation time of thestimulus animal during consecutive trials using thesame stimulus animal (habituation). On a fifth trial,a novel stimulus animal is introduced and theinvestigation time is restored to the original highestlevel (dishabituation). The second part of theparadigm, performed 24 h later, is a socialdiscrimination test. In this the test animal is given abinary choice between a previously encounteredfamiliar and a novel conspecific [22]. A specificrecognition memory, assessed by comparing thedifferential amount of time spent investigating eachstimulus animal, is considered to have been formedwhen the novel animal is investigated more thanthe unfamiliar one. This second part of the protocola l lows the recogni t ion memory tes t to be

549REVIEW: ENCODING OF OLFACTORY RECOGNITION MEMORY

performed for the familiar versus the unfamiliarstimulus animal in a single trial. This habituation/discrimination social recognition paradigmtherefore allows testing both short- and long-termmemory with a single experiment.

A second test we used as a model of socialolfactory memory in mice is the social transmissionof food preference (STFP) paradigm, which uses acombination of biological and non-biologicalodours in a social and ethologically relevantcontext. This ability of socially transmitting foodpreferences has been shown in different rodentspecies such as rats [7], mice (Mus musculus, [23, 24]and gerbils (Meriones unguiculatus [25]. Duringsocial contact between a recently fed animal(demonstrator) and a naïve conspecific (observer),olfactory cues pass from demonstrator to observersubsequently increasing the observer’s preferencefor the food its demonstrator ate [7, 26]. Observeranimals are then presented with a binary choice oftwo equally novel foods. Learning is demonstratedas an increase in probability that the observer willselect the same food as that eaten by demonstrator,over other foods [7]. A single brief interactionbetween two mice or rats is sufficient to induce afood preference that lasts for more than 6 days.This olfactory learning is believed to involve anassociation between two odours present on thedemonstrator’s breath; the odour of the recentlyeaten food and a natural constituent of rat’s breath,carbon disulfide (CS2). The association between thefood odour and the carbon disulfide is bothne c es sa ry a nd su f f i c i en t t o su p p o r t t h edevelopment of a food preference [27, 28].

All three behavioural tasks, offspring and socialrecognition and STFP are dependent on the mainolfactory system and represent good models forstudying memory formation in the main olfactorysystem. In sheep, sectioning the vomeronasal nerveorgan does not disrupt either lamb recognition ormaternal behaviour, whereas interfering with theolfactory epithelium does [29]. In rodents, lesionsin the olfactory bulb impair basic recognitionresponses in male rats [30, 31]; and chemicallyinduced anosmia can also block individualrecognition [32, 33].

The Main Olfactory System (MOS)The mammalian olfactory system is organised in

such a way that it is capable of processinginformation from two different types of stimuli:

one, a liquid-borne chemical substance known aspheromones; the second, a vast number of airbornechemicals with a large variety of structures. This isdone by segregating odour and pheromonedetection using different sensory neurones [34] andneural pathways [35] in the brain. In the mainolfactory system, odours are detected by olfactoryreceptors (OR) in the olfactory epithelium (OE) ofthe nasal cavity where every odour has a uniquecombination of responses from up to one thousanddifferent receptor types [34]. These signals are thenrelayed to the primary processing region for smell,the main olfactory bulb (MOB) and thence tosecondary and tertiary projection areas in thecortex, limbic system and thalamus (see Fig. 1).Odour signals therefore ultimately reach highercort ical areas involved in their consciousperception (attention and awareness), as well aslimbic areas, such as the amygdala and thehypothalamus, that are involved in emotional andmotivational responses.

In brief, mitral and tufted cells of the MOB (c.f.below for detailed description of the MOB) sendtheir axons, via the lateral olfactory tract (LOT), tothe outer layer of the olfactory cortex [36]–thepiriform cortex, the olfactory tubercle, anteriorolfactory nucleus and specif ic parts of theentorhinal cortex and amygdala [37–39]. From theolfactory cortex, there is a substantial projection tothe mediodorsal thalamus and the ventral part ofthe submedius nucleus (SM), which in turn projectsto the frontal cortex [40]. Disruption of theseconnections results in deficits in complex but not insimple odour discrimination tasks [41–43].Another small projection from the LOT goes to thelateral entorhinal cortex [44] and thence via theperforant path to the dentate gyrus and CA1/CA3of the hippocampus. The olfactory cortex alsoprojects to the hypothalamus and the anteriorcortical amygdala [45]. This connection could bethe one responsible for the regulation of several ofthe autonomic and neuroendocrine changes thatoccur in response to odours [46].

Learning in the MOSOdour learning in the main olfactory system,

unlike that involving the accessory olfactorysystem [48], is thought to be more distributed, withsecondary and tertiary olfactory processing regionsbeing involved. The activation and involvement ofthese brain areas depends to some extent, on the


nature of the task being performed and on temporalconfigurations. Dynamic changes occur in bothneural substrates and the molecular involvement information, short- and long-term memory traces.Studies of brain structure activation and reversibleinactivation of some key structures have providedsome valuable information. Initial memoryformation and its short-term retention seem torequire a distributed neural system, involving theolfactory bulb, piriform and entorhinal cortices andthe hippocampus. Our early work aimed todifferentiate neural substrates controlling maternalbehaviour from those involved in olfactorymemory formation, by comparing post-parturientewes exposed to their lambs for 30 min tohormonally primed animals that received VCS andwere not exposed to lambs. This showed thatolfactory memory formation involved activation,quantified by measurement of altered mRNAexpression of the immediately-early genes c-fosand/or zif/268, of the MOB, of piriform, entorhinaland orbitofrontal cortices and dentate gyrus [49].Studies comparing the response of intact andanosmic ewes to their own lamb have also shownthat anosmic ewes, displaying maternal behaviourbut not individual lamb recognition, fail to showincreased expression in the MOB, piriform andfrontal cortices and cortical amygdala [50, 51]. Forsubsequent memory consolidation, 4.5 h post-partum, the piriform and entorhinal corticesseemed to play a key role, as shown by a studyusing expression of the brain derived neurotrophic

factor (BDNF) and its receptor tyrosine receptorkinase B (trk-B)–a measure of neural plasticity [52].We recently carried out a systematic study of thetemporal involvement of the major areas in theolfactory system, by temporally inactivating them(bilaterally) with either a local anaesthetic( te t raca ine) or a GABA a receptor agonis t(muscimol) directly infused via microdialysisprobes. This showed that the olfactory bulb, thepiriform and the entorhinal cortices are essential forthe formation of a normal selective recognitionmemory. However, complete disruption, was onlyseen with the olfactory bulb and the entorhinalcortex [53].

There have been fewer studies regarding theinvolvement of neural substrates in socialrecognition in mice. Studies using oxytocinreceptor knockout mice have shown that a briefsocial exposure leads to regional activation of theolfactory system as evidenced by increased c-fOS

expression in the olfactory bulb, piriform cortexand the medial amygdala [54]. Non-behaviouralstudies have also shown plasticity changes in boththe OB and piriform cortex resulting from OBstimulation. High frequency stimulation of thegranule cell layer of the OB results in a selectivelong-term potentiation in both the OB itself and thepiriform cortex [55]. Recently we have also shownthat consecutive 15 min periods of NMDA-stimulation to the MOB result in a long termpotentiation of NMDA-evoked neurotransmitterrelease in both the olfactory bulb and the piriform

Fig. 1. Schematic diagram showing the major connections of the main olfactory system (not all projections areshown, nor the extensive reciprocal projections between olfactory areas). Olfactory memory involveschanges occurring within the olfactory bulb as well as in its secondary (piriform cortex) and tertiaryprojection sites (entorhinal and orbitofrontal cortices and the hippocampus). Adapted from [47].


cortex after 4 h [4, 5].Fo l lowing memory consol idat ion , a f ter

approximately 8 h, only the olfactory bulb andpiriform cortex appear to be important for recall.For olfactory memory recall, there seems to bedistinct neural processes for short- and long-termmemory. We have found that for ewes that hadformed an o l fac tory memory of i t s lamb,inactivation of the entorhinal cortex interfered withits recall in the short-term (4 h post-memoryformation) but not in the long-term (12 h to 6 dayspost-memory formation). However, inactivation ofthe olfactory bulb disrupted recall at both timepoints [53]. Consistent with this, it has been shownthat the piriform cortex is activated by recognitionof the lamb after both 4 h and 7 days contactdurations [50]. In rodents, lesions of the entorhinalcortex result in deficits of short-term odourmemory [56–58]. Other studies using a modifiedversion of the Coolidge effect, another socialolfactory learning test, have also shown that malehamsters with ibotenic acid lesions of theperirhinal-entorhinal cortex have impaired short-term social discrimination memory. Lesions to thehippocampus, however, have no effect on short-term recognition memory [59].

Olfactory recognition memory, therefore,requires a more distributed neural network in theolfactory system for memory formation and short-term memory, with the entorhinal cortex playing akey role. As the memory consolidates, this networkis reduced, and it seems that only the OB andpiriform cortex are needed to maintain a memorytrace throughout its existence.

The Main Olfactory Bulb (MOB)As already discussed, many studies have

consistently reported that a general characteristic ofolfactory memory is that it induces long-lastingneural changes at the level of the olfactory bulb.The main olfactory bulb (MOB) has a simplestructure involving three main layered cell types(Fig. 2): periglomerular, mitral/tufted and granulecells. The olfactory nerve terminals synapse withmitral and tufted cells in the MOB glomerulus, andalso with intrinsic periglomerular cells (PG) thatcircumscribe the glomeruli. PG cells (GABAergicor dopaminergic) mediate feedback inhibitionwithin the glomerulus and lateral (feedforward)inhibition between glomeruli [60–62]. Mitral andtufted cells send their primary dendrites into the

glomeruli and their secondary dendrites into theexternal plexiform layer. Mitral cells use glutamateand possibly aspartate as transmitters and formreciprocal dendro-dendritic synapses with GABA-containing inhibitory interneurones at distinctlevels. Granule cells, which are the most numerousneurones in the olfactory bulb, in turn, makesynaptic contacts with the lateral secondarydendrites of the mitral cells in the externalplexiform layer and are likely to be involved in thecontrol of mitral cell output [63, 64].

Mechanisms of Olfactory Learning In early postpartum period (2–4 h) a maternal

ewe’s selective recognition of her lamb can beshown by physiological changes in the olfactorybulb that parallel observable behavioural changes.Electrophysiological recordings in the mitral celllayer of the olfactory bulb have shown that severald a y s p o s t p a r t u m 6 0 % o f c e l l s r e s p o n dpreferentially to lamb’s odour compared toprepartum period where no preferential activitywa s seen [65] . Of the ce l l s that respondsignificantly to these lamb odours the majority(70%) do so equivalently to own or strange lambs,however 30% did respond preferentially to theodours of the ewe’s own lamb. In addition,a l t h o u g h l a m b o d o u rs d o n o t s t i m u l a t eneurotransmitter release in the olfactory bulbprepartum, by 4 h postpartum, all lamb odoursi n c r e a s e c e n t r i f u g a l t r a n s m i t t e r r e l e a s e(noradrenaline (NA) and acetylcholine (ACh))while own lamb odours selectively increaseintrinsic glutamate and GABA release [12].Associated with these neurochemical changes isevidence for enhanced glutamatergic stimulation ofGABA release from the granule cells followinglearning [12].

How does the reorganisation of the olfactorybulb take place? Both centrifugal and intrinsicolfactory bulb pathways play an important role inestablishing the plasticity changes in the olfactorybulb that lead to olfactory memory formation.Lesions of noradrenergic projections to theolfactory bulb, or direct infusion of β-noradrenergicantagonists, significantly reduce the number ofanimals developing the olfactory recognitionmemory [66]. Similar effects are seen followingtreatments with cholinergic antagonists such asscopolamine [67]. The result of the changes in NAand ACh in the olfactory bulb is to reduce


GABAergic feedback inhibition from the granulecells to the mitral cells. The disinhibited mitral cellsbecome more active in their response to odours andthrough a glutamate–AMPA/NMDA receptor–nitric oxide–cyclic GMP signalling cascade there isupregulation both of mitral to granule cell synapsesas well as of auto-excitation synapses on the mitralcells themselves. Thus pharmacological inhibitionof NMDA/AMPA receptors, nitric oxide synthase(NOS) or soluble guanylate cyclase all prevent theformation of an olfactory reccogition memory andthe ability of the lamb odours to evoke enhancedglutamate and GABA recelease [68]. However oncea memory has been allowed to form similarinhibition does not interfere with recall orpotent ia ted g lutamate and GABA re leasesuggesting that these signalling pathways are notinvolved in recall of a consolidated recognitionmemory [61].

Thus the overall signalling pathway model wesuggest is involved in olfactory memory formationin the olfactory bulb [3, 68] is as follows (Fig. 2):Noradrenaline (or acetylcholine) release (triggeredby attention, arousal or vaginocervical stimulation)at the centrifugal projection synapses withGABAergic granule interneurones reduces GABA

release onto mitral cells and consequentlydisinhibits them [12]. With incoming odour-induced activation of olfactory receptors in thepresence of this disinhibited state, mitral cellsincrease their release of glutamate at theirreciprocal synapses [69] activating an intracellularmolecular cascade that will ultimately release NOfrom the granule cells. In turn, NO acts as aretrograde messenger to stimulate guanylyl cyclaseactivity and the formation of cyclic GMP in the pre-synaptic mitral cell, which then facilitates furtherglutamate release. This glutamate release promotesenhanced GABA release from the granule cellsthereby increasing feedback inhibition onto themitral cells. It also facilitates autoreceptors on thepresynaptic site so that there is enhanced glutamaterelease following exposure to the learned odours.In this way learning causes increased sensitivity toglutamate in both mitral cells and at the mitral-to-granule cell synapses. This results in bothincreased excitability of the mitral cells and tighterexcitatory/inhibitory coupling between them andtheir associated granule cells such that learnedodours evoke both a stronger activation of themitral cells which are tuned to it and a sharpeningof their phasic firing discharge pattern caused by a

Fig. 2. Left: Schematic representation of cell organisation in the main olfactory bulb (MOB), showing inputs(olfactory receptors, centrifugal pathways) and outputs (mitral cells) to the MOB and its intrinsicorganisation and neurotransmitters. Right: Memory formation is associated with enhanced mitral-to-granule and periglomerular cell synapses to bias the network to respond to the learned odour. Thisinvolves both, increased acetylcholine (ACh) or noradrenaline (NA) from the centrifugal projections ,that in turn reduce GABA release onto mitral cells and glutamate-evoked nitric oxide/cGMP releasepotentiating glutamate and subsequent GABA release from granule and periglomerular cells. GLU,glutamate; L-arg, L-arginine; L-cit, L-citrulline; nNOS, neuronal nitric oxide synthase; NO, nitric oxide;NMDA-R, NMDA receptor; AMPA/Kainate-R, AMPA/Kainate receptor; mGluR, metabotropic glutamatereceptor.


more robust inhibitory feedback response from thegranule cells that follows from each excitatory burst[68, 70].

This process is not unique to the postpartumewe, as similar changes have been observed in micewith olfactory conditioning [70], social recognitionand STFP [4]. In our model of social olfactorymemory using the social recognition habituation/discrimination test and the STFP test, long-term (24h) memory formation can be prevented by blockingei ther NMDA (using the non-compet i t ivea n t a g o n i s t M K 8 0 1 ) o r A M P A ( u s i n g t h ecompetitive antagonist NBQX) receptor activation(Fig. 3). Similarly, preventing NO release byinhibiting NOS (using the non-selective NOSinhibitor, L-NARG) impaired formation of long-term olfactory memory in both models [4]. With

these olfactory learning models too disruption ofNMDA or NO signalling after memory acquisitionfailed to have any influence on subsequent recall[5]. Thus, as with the sheep model, glutamateacting on ionotropic glutamate receptors andpromoting NO release is required for the initialacquisition phase of an olfactory recognitionmemory trace but not its recall post-acquisition.

In both sheep and mouse models of olfactoryinduced learning or plasticity there is evidence ofaltered sensit ivity of class 1 metabotropicglutamate (mGluR) receptors in mitral and granulecells and agonists of this class of receptor are morepotently able to release glutamate and GABAwithin the olfactory bulb following consolidation ofl e a r n i n g i n s h e e p [ 4 8 , 6 4 ] . P r e l i m i n a r ypharmacological studies have also shown that

Fig. 3. Effects of the AMPA and NMDA receptor antagonists, NBQX (left panel) and MK801 (centre panels) respectively, andNOS inhibitor L-NARG (right panels) on acquisition of social olfactory memories, tested at 24 h with the habituation/discrimination social recognition task (top panels) and STFP (bottom panels) task. Control mice, injected with salinesolution (and/or the inactive enantiomer D-NARG for the L-NARG experiment), showed both a discriminationbetween familiar and unfamiliar stimulus mice (paired t-test, P<0.05, P<0.01 and P<0.001 vs unfamiliar) and preferencefor the cued food (paired t-test, * P<0.01 vs cued food). All three treatments resulted in impaired 24 h memory retentionin both tasks of social olfactory memory. Taken from [4].


olfactory bulb infusions of antagonists targetingmGluR1 (AIDA and LY393675) but not mGluR5(MPEP) prevent recall of the recognition memoryafter it is consolidated (12–24 h post-partum) buthave no impact on memory formation [53, 68, 70].Therefore, as the olfactory recognition memoryconsolidates there is a shift away from thedependence on glutamate acting on NMDA/AMPA-NO signalling pathways to a dependenceon its actions on mGluR1 receptors. This effectivelyreleases the former signalling pathway to beutilised for formation of new memory associations.

Olfactory Learning and Ovarian HormonesThe sheep offspring recognition model is based

on a biologically distinct event which occurs duringa short time-window where the ewe is primedappropriately with oestrogen and progesteroneduring pregnancy and experiences vaginocervicalstimulation (VCS) as a result of giving birth.Olfactory memory can even be induced by artificialVCS but only after ewes have been treated witheither oestrogen or a combination of oestrogen andprogesterone [13]. This strongly suggests that sexhormones play a key role in olfactory memoryformation processes described above. However,little has been done to establish the precise effectsthat gonadal hormones may have on the olfactorysystem and olfactory learning. In male rats,castration reduces NA concentrations in the OB,but not in the olfactory cortex [71, 72], and reducespotassium-evoked release of NA in the OB [71, 73].It also substantially impairs olfactory socialrecognition [74] and systemic treatment of castratedmales with L-DOPA restores social recognition topre-castration levels [75]. Very little work has beencarried out to establish effects of ovarian hormoneson the olfactory system. The first systematic studyrelating ovarian hormones to social recognitionmemory in female rats [76] showed a memory-retention improvement effect of oestradiol.Ovariectomized (OVX) females had a 30 min, butnot 120 min memory retention for the familiar rat,contrasting with oestradiol-treated OVX (OVX+E2)animals that recognised the familiar animal at both30 and 120 min time points. Memory retentionimpairment did not reappear until 6 weeks aftertermination of oestradiol treatment, indicating thatlong-term effects of hormones may be involved. Amore recent study has shown similar results thatare dependent on the use of high doses of oestrogen

treatment. Ovariectomized mice receiving a four-week oestrogen-treatment at a high dose (0.72 mgof 17β-oestradiol) showed a long-term (24 h) socialrecognition memory whereas OVX and low doseoestradiol treated animals did not [77]. This makesit important to determine whether hormonalfluctuations in the normal physiological range canhave any impact on olfactory memory function,and one obvious place to start is by studyingchanges during the course of natural oestrouscycles.

Our early work studied the effect of hormonalfluctuations during the rat oestrous cycle on OBactivity evoked by vaginocervical stimulation(VCS). Such VCS resulting from parturition (ormating) enhances NA input to the OB, necessaryfor memory formation. In vivo microdialysisexperiments have shown no oestrous cycle effectson basal concentrations of classical transmitters andnitric oxide (NO) in the OB. However, VCSsignificantly increases concentrations of glutamate,aspartate, GABA, noradrenaline, dopamine andnitric oxide (NO) in females during proestrous/oestrous, but not during metaoestrous/dioestrousor following ovariectomy [78]. This suggests OBmitral cells are excited by VCS only when femalesare at proestrous/oestrous stage. Similar resultsw er e o b t a i ne d u s i n g p o t a ss i u m e v o k e dneurochemical release in the olfactory bulb,confirming that sex hormones changes can indeedact directly at the level of the OB and may beimportant for mediating memory related plasticitychanges.

We decided to test this oestrous cycle effect at abehavioural level using the two tests of socialolfactory memory in mice, the habituation/discrimination and the STFP task. In both tests,significant improvement of learning retentionoccurs when original learning takes place duringthe proestrus phase of the ovarian cycle [5]. Nodifferences on either olfactory perception ormotivation were seen across the oestrous cycle (Fig.4). Our results suggest that olfactory memoryformation is facilitated in females when they arereproductively active. This might be because whenfemales are at a time when they are most likely toconceive they need to assess the environment forpresent dangers and for i ts suitabil i ty forreproductive activities (e.g. sex behaviour,pregnancy, and offspring rearing). They must beprepared to either engage in sex behaviour when


the environment predicts success–familiar socialenvironment will improve reproductive andrearing success–, or suspend resource-costlyreproductive behaviours until they are more likelyto pay off [79].

Recent studies reporting that female mice lackingfunctional oestrogen receptors (ERα, and ERβ) failto form social memories, suggest that the ovariancycle effects may be mediated via oestrogen actingon these two oestrogen receptors [80]. Sinceoestrogen receptors have been found in theolfactory bulb [81], we have recently investigatedthe impact of the oestrous cycle on the NMDA/NOsignalling pathway that is critical for olfactorymemory formation. Results have shown that

ovarian hormone actions on olfactory memory maybe via modulation of NO availability since animalslacking the functional expression of neuronal nitricoxide synthase (nNOS–/–) do not show oestrouscycle dependant effects on memory retention. Onthe other hand, exogenous treatment with the NOdonor molsidomine, improves memory retention indioestrous and oestrous females to the levels seenduring proestrous [5]. It would therefore appearthat oestrogen may be influencing memoryformation in the olfactory bulb both throughmodulation of NA and the subsequently ability ofglutamate to stimulate NO release. However theprecise mechanisms of action still require furtherinvestigation.

References

1. Ferguson JN, Young LJ, Hearn EF, Matzuk MM,Insel TR, Winslow JT. Social amnesia in micelacking the oxytocin gene. Nat Genet 2000; 25: 284–288.

2. Brennan PA, Kendrick KM, Keverne EB.Neurotransmitter release in the accessory olfactorybulb during and after the formation of an olfactorymemory in mice. Neuroscience 1995; 69: 1075–1086.

3. Kendrick KM, Da Costa AP, Broad KD, Ohkura S,Guevara R, Lery F, Keverne EB. Neural control of

maternal behaviour and olfactory recognition ofoffspring. Brain Res Bull 1997; 44: 383–395.

4. James B. The importance of NMDA/nitric oxidesignalling in mouse olfactory memory. In:Laboratory of Cognitive & BehaviouralNeuroscience, Babraham Institute; 2003; Universityof Cambridge, Cambridge. 230.

5. Sánchez-Andrade G. Ovarian cycle influences onneural plasticity and olfactory learning in themouse. In: Laboratory of Cognitive & Behavioural

Fig. 4. Effect of oestrous cycle on acquisition of social olfactory memories, tested at 24 h with thehabituation/discrimination social recognition task (left panels) and STFP (right panels) task. Femalesshowed both a social recognition ( P<0.05, vs unfamiliar) and a food preference (*** P<0.001, vscued food) at 24 h, only when learning occurred during the proestrous stage of the oestrous cycle.Females during oestrous and dioestrous showed a significantly impaired 24 h olfactory memoryretention when compared to females during proestrous (+ P<0.05, ++ P<0.01). There was nodifference across the oestrous cycle on the habituation-dishabituation response, all females showed asignificant TRIAL effect (P<0.001). Taken from [5].


Neuroscience, Babraham Institute; 2005; Universityof Cambridge, Cambridge. 271.

6. Burne THJ, Wilkinson LS, Keverne BE, KendrickKM. Effects of anaesthetic agents on sociallytransmitted olfactory memories in mice. In: SeventhConference on the Neurobiology of Learning andMemory; 2001; Irvine, CA, USA.

7. Galef BG, Wigmor SW. Transfer of informationconcerning distant foods: a laboratory investigationof the ‘information-centre’ hypothesis. Anim Behav1983; 31: 748–758.

8. Lai WS, Johnston RE. Individual recognition afterfighting by golden hamsters: a new method. PhysiolBehav 2002; 76: 225–239.

9. Insel TR, Young LJ. The neurobiology ofattachment. Nat Rev Neurosci 2001; 2: 129–136.

10. Johnston RE, Rasmussen K. Individual recognitionof female hamsters by males–role of chemical cuesand of the olfactory and vomeronasal systems.Physiol Behav 1984; 33: 95–104.

11. Carr WJ, Hirsch JT, Balazs JM. Responses of male-rats to odors from familiar vs novel females. BehavNeural Biol 1980; 29: 331–337.

12. Kendrick KM, Levy F, Keverne EB. Changes in thesensory processing of olfactory signals induced bybirth in sheep. Science 1992; 256: 833–836.

13. Kendrick KM, Levy F, Keverne EB. Importance ofvaginocervical stimulation for the formation ofmaternal bonding in primiparous and multiparousparturient ewes. Physiol Behav 1991; 50: 595–600.

14. Carr WF, Yee L, Gable D, Marasco E. Olfactoryrecognition of conspecifics by domestic Norwayrats. J Comp Physiol Psycol 1976; 90: 821–828.

15. Sawyer TF, Hengehold AK, Perez WA.Chemosensory and hormonal mediation of socialmemory in male rats. Behav Neurosci 1984; 98: 908–913.

16. Thor DH, Holloway WR. Social memory of themale laboratory rat. J Comp Physiol Psychol 1982; 96:1000–1006.

17. Axelson JF, Smith M, Duarte M. Prenatal flutamidetreatment eliminates the adult male rat’sdependency upon vasopressin when forming social-olfactory memories. Horm Behav 1999; 36: 109–118.

18. Bluthe RM, Dantzer R. Social recognition does notinvolve vasopressinergic neurotransmission infemale rats. Brain Res 1990; 535: 301–304.

19. Kogan JH, Frankland PW, Silva AJ. Long-termmemory underlying hippocampus-dependent socialrecognition in mice. Hippocampus 2000; 10: 45–56.

20. Dluzen DE, Kreutzberg JD. 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (Mptp) disrupts socialmemory recognition processes in the male mouse.Brain Res 1993; 609: 98–102.

21. Winslow JT, Camacho F. Cholinergic modulation ofa decrement in social-investigation followingrepeated contracts between mice.

Psychopharmacology 1995; 121: 164–172.22. Engelmann M, Wotjak CT, Landgraf R. Social

discrimination procedure: an alternative method toinvestigate juvenile recognition abilities in rats.Physiol Behav 1995; 58: 315–321.

23. Valsecchi P, Galef BG. Social influence on the foodpreference of house mice (Mus Musculus). Int J CompPsychol 1989; 2: 245–256.

24. Gass P, Wolfer DP, Balschun D, Rudolph D, FreyU, Lipp HP, Schutz G. Deficits in memory tasks ofmice with CREB mutations depend on gene dosage.Learn Mem 1998; 5: 274–288.

25. Valsecchi P, Choleris E, Moles, A, Guo C,Mainaardi M. Kinship and familiarity as factorsaffecting social transfer of food preferences in adultMongolian gerbils (Meriones unguiculatus). J CompPsychol 1996; 110: 243–251.

26. Strupp BJ, Levitsky DA. Social transmission of foodpreference in adult Hooded rats (rattus norvegicus). JComp Psychol 1984; 98: 257–266.

27. Bean NJ, Galef BG, Mason JR. The effect of carbondisulfide on food consumption by house mice. JWildlife Manag 1988; 52: 502–507.

28. Galef BG Jr, Mason JR, Preti G, Bean NJ. Carbondisulfide: a semiochemical mediating socially-induced diet choice in rats. Physiol Behav 1988; 42:119–124.

29. Levy F, Locattelli A, Piketty V, Tillet Y, PoindronP. Involvement of the main but not the accessoryolfactory system in maternal behavior ofprimiparous and multiparous ewes. Physiol Behav1995; 57: 97–104.

30. Dantzer R, Tazi A, Bluthe RM. Cerebrallateralization of olfactory-mediated affectiveprocesses in rats. Behav Brain Res 1990; 40: 53–60.

31. Bluthe RM, Dantzer R. Chronic intracerebralinfusion of vassopressin antagonist modulatessocial recognition in rat. Brain Res 1992; 572: 261–264.

32. Matochick JA. Role of the main olfactory system inrecognition of between individual spiny mice.Physiol Behav 1998; 42: 217–222.

33. Popik P, Vetulani J, Bisaga A, van Ree JM.Recognition cue in the rat's social memoryparadigm. J Basic Clin Physiol Pharmacol 1991; 2: 315–327.

34. Buck LB. The molecular architecture of odor andpheromone sensing in mammals. Cell 2000; 100: 611–318.

35. Shipley MT, Adamek GD. The connections of themouse olfactory bulb: a study using orthograde andretrograde trensport of wheat germ agglutininconjugated to horseradish peroxidase. Brain Res Bull1984; 12: 669–688.

36. Price JL. Autoradiographic study of complementarylaminar patterns of termination of afferent fibers toolfactory cortex. J Comp Neurol 1973; 150: 87–108.


37. Scott JW. The olfactory-bulb and central pathways.Experientia 1986; 42: 223–232.

38. Shipley MT, Ennis M. Functional organization ofolfactory system. J Neurobiol 1996; 30: 123–176.

39. Scott JW. Organization of olfactory-bulb outputcells and their local circuits. Ann N Y Acad Sci 1987;510: 44–48.

40. Price JL, Slotnick BM. Dual olfactoryrepresentation in the rat thalamus–an anatomicaland electro-physiological study. J Comp Neurol 1983;215: 63–77.

41. Zhang YP, Burk JA, Glode BM, Mair RG. Effects ofthalamic and olfactory cortical lesions oncontinuous olfactory delayed nonmatching-to-sample and olfactory discrimination in rats (Rattusnorvegicus). Behav Neurosci 1998; 112: 39–53.

42. McBride SA, Slotnick B. The olfactorythalamocortical system and odor reversal learningexamined using an asymmetrical lesion paradigm inrats. Behav Neurosci 1997; 111: 1273–1284.

43. Koger SM, Mair RG. Comparison of the effects offrontal cortical and thalamic lesions on measures ofolfactory learning and memory in the rat. BehavNeurosci 1994; 108: 1088–1100.

44. Krettek JE, Price JL. Projections from theamygdaloid complex and adjacent olfactorystructures to the entorhinal cortex and to thesubiculum in the rat and cat. J Comp Neurol 1977;172: 723–752.

45. Price JL, Slotnick BM, Revial MF. Olfactoryprojections to the hypothalamus. J Comp Neurol1991; 306: 447–461.

46. McClintock MK. Estrous synchrony and itsmediation by airborne chemical communication(rattus-norvegicus). Horm Behav 1978; 10: 264–276.

47. Brennan PA, Keverne EB. Neural mechanisms ofmammalian olfactory learning. Prog Neurobiol 1997;51: 457–481.

48. Brennan PA, Kaba H, Keverne EB. Olfactoryrecognition: a simple memory system. Science 1990;250: 1223–1226.

49. Da Costa APC, Broad KD, Kendrick KM. Olfactorymemory and maternal behaviour-induced changesin c-fos and zif/268 mRNA expression in the sheepbrain. Mol Brain Res 1997; 46: 63–76.

50. Keller M, Merisse M, Jouaneau N, Venier G,Archer E, Levy F. Neural networks involved inacquisition and consolidation of the lamb odour byparturient ewes. In: 35th Annual General Meeting ofthe Brain and Behaviour Society; 2003.

51. Keller M, Meurisse M, Levy F. Mapping the neuralsubstrates involved in maternal responsiveness andlamb olfactory memory in parturient ewes using Fosimaging. Behav Neurosci 2004; 118: 1274–1284.

52. Broad KD, Mimmack ML, Keverne EB, KendrickKM. Increased BDNF and trk-B mRNA expressionin cortical and limbic regions following formation of

a social recognition memory. Eur J Neurosci 2002; 16:2166–2174.

53. Insel TR, Laurent G, Kendrick KM, Young LJ,Eichenbaum H. Olfactory memory: how does thenose know. In: Society for Neuroscience 31st AnnualMeeting; 2001; San Diego, CA, USA.

54. Ferguson JN, Aldag JM, Insel TR, Young LJ.Oxytocin in the medial amygdala is essential forsocial recognition in the mouse. J Neurosci 2001; 21:8278–8285.

55. Patneau DK, Stripling JS. Functional correlates ofselective long-term potentiation in the olfactorycortex and olfactory bulb. Brain Res 1992; 585: 219–228.

56. Staubli U, Ivy G, Lynch G. Hippocampaldenervation causes rapid forgetting of olfactoryinformation in rats. Proc Nat Acad Sci USA 1984; 81:5885–5887.

57. Kaut KP, Bunsey MD. The effects of lesions to therat hippocampus or rhinal cortex on olfactory andspatial memory: retrograde and anterogradefindings. Cognit Affect Behav Neurosci 2001; 1: 270–286.

58. Kaut KP, Bunsey MD, Riccio DC. Olfactorylearning and memory impairments followinglesions to the hippocampus and perirhinal-entorhinal cortex. Behav Neurosci 2003; 117: 304–319.

59. Petrulis A, Eichenbaum H. The perirhinal-entorhinal cortex, but not the hippocampus, iscritical for expression of individual recognition inthe context of the Coolidge effect. Neuroscience 2003;122: 599–607.

60. Shepherd GM. Synaptic organization ofmammalian olfactory bulb. Physiol Rev 1972; 52:864–917.

61. Gall CM, Hendry SHC, Seroogy KB, Jones EG,Haycock JW. Evidence for coexistence of GABA anddopamine in neurons of the rat olfactory-bulb. JComp Neurol 1987; 266: 307–318.

62. Kosaka T, Hataguchi Y, Hama K, Nagatsu I, Wu JY.Coexistence of immunoreactivities for glutamate-decarboxylase and tyrosine-hydroxylase in someneurons in the periglomerular region of the rat mainolfactory bulb—possible coexistence of gamma-aminobutyric acid (GABA) and dopamine. Brain Res1985; 343: 166–171.

63. Brennan PA, Friedrich RW. Something in the air:new vistas on olfaction. Trends Neurosci 1998; 21: 1–2.

64. Trombley PQ, Shepherd GM. Synaptictransmission and modulation in the olfactory bulb.Curr Opin Neurobiol 1993; 3: 540–547.

65. Kendrick KM. Neurobiological correlates of visualand olfactory recognition in sheep. Behav Process1994; 33: 89–111.

66. Levy F. Neurobiological mechanisms involved inrecognition of olfactory signature of the young in


sheep. J Sociét Biologie 2002; 196: 77–83.67. Levy F, Richard P. Meurisse M, Ravel N.

Scopolamine impairs the ability of parturient ewesto learn to recognise their lambs. Psychopharmacology1997; 129: 85–90.

68. Kendrick KM, Guevara-Guzman R, Zorrila J,Hinton MR, Broad KD, Mimmack M, Ohkura S.Formation of olfactory memories mediated by nitricoxide. Nature 1997; 388: 670–674.

69. Trombley PQ. Noradrenergic modulation ofsynaptic transmission between olfactory bulbneurons in culture—implications to olfactorylearning. Brain Res Bull 1994; 35: 473–484.

70. Brennan PA, Schellinck HM, De la Riva C,Kendrick KM, Keverne EB. Changes inneurotransmitter release in the main olfactory bulbfollowing an olfactory conditioning procedure inmice. Neuroscience 1998; 87: 583–590.

71. Cornwell-Jones CA, Marasco EM. Castrationdecreases olfactory bulb norepinephrine in male ratsbut not in hamsters. Brain Res 1980; 183: 377–382.

72. Shang YL, Dluzen DE. Castration reduces olfactorybulb norepinephrin transporter function asindicated by response to noradrenergic uptakeblockers. Brain Res 1998; 779: 119–124.

73. Guan XB, Dluzen D. Castration reduces potassium-stimulated norepinephrine release from superfusedolfactory bulbs of male rats. Brain Res 1991; 568: 147–151.

74. Guan X, Blank J, Dluzen D. Depletion of olfactorybulb norepinephrine by 6-OHDA disrupts chemicalcue but not social recognition responses in male rats.Brain Res 1993; 622: 51–57.

75. Shang Y, Dluzen DE. Nisoxetine infusion into the

olfactory bulb enhances the capacity for male rats toidentify conspecifics. Neuroscience 2001; 104: 957–964.

76. Hlinak Z. Social recognition in ovariectomized andestradiol-treated female rats. Horm Behav 1993; 27:159–166.

77. Nakazawa M, Tang AC, Reeb BC, Sisti HM,Romeo RD, McEwen BS. Effects of estrogenreplacement on Hypothalamic-Pituitary-Adrenal(HPA) axis, social recognition, and open fieldbehaviour. In: Society for Neuroscience 33rd AnnualMeeting; 2003; New Orleans, LA, USA.

78. Guevara-Guzman R, Barrera-Mera B, De la riva C,Kendrick KM. Release of classical transmitters andnitric oxide in the rat olfactory bulb, evoked byvaginocervical stimulation and potassium, varieswith the oestrous cycle. Eur J Neurosci 2000; 12: 80–88.

79. Morgan MA, Schulkin J, Pfaff DW. Estrogens andnon-reproductive behaviors related to activity andfear. Neurosci Biobehav Rev 2004; 28: 55–63.

80. Choleris E, Gustafsson JA, Korach KS, Muglia LJ,Pfaff DW, Ogawa S. An estrogen-dependent four-gene micronet regulating social recognition: a studywith oxytocin and estrogen receptor-alpha and -betaknockout mice. Proc Nat Acad Sci USA 2003; 100:6192–6197.

81. Mitra SW, Hoskin E, Yudkovitz J, Pear L,Wilkinson HA, Hayashi S, Pfaff DW, Ogawa S,Rohrer SP, Schaeffer JM, McEwen BS, Alves SE.Immunolocalization of estrogen receptor beta in themouse brain: comparison with estrogen receptoralpha. Endocrinology 2003; 144: 2055–2067.

neural encoding of olfactory recognition memory › 6d14 › b1d4cb3ab7dd... · neural encoding of...

Documents