Olfaction, Emotion & the Amygdala: arousal-dependent modulation of

20 Pages·2004·0.5 MB·English

Impulse: The Premier Journal for Undergraduate Publications in the Neurosciences
June 2004, 1 (1): 1-58
Olfaction, Emotion & the Amygdala: arousal-dependent
modulation of long-term autobiographical memory and its
association with olfaction: beginning to unravel the Proust
phenomenon?
Mark Hughes1
1University of Edinburgh, Centre for Neuroscience, No. 1 George Square, Edinburgh, UK
The sense of smell is set apart from other very old autobiographical memories
sensory modalities. Odours possess the (episodes known anecdotally as Proust
capacity to trigger immediately strong phenomena). Particular emphasis is placed
emotional memories. Moreover, odorous on the diversity of functions attributed to
stimuli provide a higher degree of memory the amygdala. It’s role in modulating the
retention than other sensory stimuli. Odour encoding and retrieval of long-term
perception, even in its most elemental memory is investigated with reference to
form - olfaction - already involves limbic lesion, electrophysiological, immediate
structures. This early involvement is not early gene, and functional imaging studies
paralleled in other sensory modalities. in both rodents and humans. Additionally,
Bearing in mind the considerable the influence of hormonal modulation and
connectivity with limbic structures, and the the adrenergic system on emotional
fact that an activation of the amygdala is memory storage is outlined. I finish by
capable of instantaneously evoking proposing a schematic of some of the
emotions and facilitating the encoding of critical neural pathways that underlie the
memories, it is unsurprising that the sense odour-associated encoding and retrieval of
of smell has its characteristic nature. emotionally toned autobiographical
The aim of this review is to analyse current memories.
understanding of higher olfactory
information processing as it relates to the
ability of odours to spontaneously cue Key Words: Olfaction; Emotion; Proust
highly vivid, affectively toned, and often phenomenon; Amygdala.
Introduction epilepsy, in states of hypnosis or psychosis,
and, less radically, in everybody in response to
the mnemonic stimulus of certain words,
“Smell is a potent wizard that transports us
sounds, scenes, and especially smells. Odours
across thousands of miles and all the years we
are rarely perceived in a purely objective and
have lived… odours, instantaneous and
neutral way, often being loaded with appealing
fleeting, cause my heart to dilate joyously or
or repugnant feelings. Olfactory pathways
contract with remembered grief.” Helen
involve anatomical structures that are also
Keller (blind and deaf educator, 1880-1968)
heavily implicated in emotion and memory.
This quotation articulates a neurophysiological
The amygdala is one such structure. The
interaction between the systems of olfaction,
challenge is to propose a model of the neural
emotion and memory. Hughlings Jackson
mechanisms that underlie encoding and
(1880), arguably the founder of neurology as a
retrieval of odour-cued emotionally toned
science, termed this kind of forced
autobiographical memories (see figure one).
reminiscence as “a doubling of
consciousness.” This occurrence is not
uncommon during attacks of migraine or
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
Proust Phenomena it was encoded results in a better memory
performance than encoding and retrieving in
different contexts (Eich et al., 1985). Schab
The ability of odours to spontaneously cue
(1990) used ambient odour as a context,
highly vivid, affectively toned, and often very
demonstrating that reinstating the same odour
old autobiographical memories is known as
at test as that present at study appears to
the Proust phenomenon (Chu & Downes,
improve memory performance in humans over
2000) and is derived from the experiences of
conditions where the ambient odour is not
Proust himself (1922):
reinstated at test. By manipulating odours that
“And soon, mechanically, weary after a dull
were appropriate or inappropriate to a
day with the prospect of a depressing morrow,
situation, Herz (1997) has also shown in
I raised to my lips a spoonful of tea in which I
humans that distinctive odours result in greater
had soaked a morsel of cake. No sooner had
context effects.
the warm liquid, and the crumbs with it,
These studies indicate that ambient odours
touched my palate than a shudder ran through
present at the time of an event can be encoded
my whole body, and I stopped, intent upon the
in parallel with event details and consequently
extraordinary changes that were taking
be used as cues in the retrieval of those event
place…I was conscious that it was connected
details. The types of memory commonly
with the taste of tea and cake, but that it
reported in connection with Proust phenomena
infinitely transcended those savours, could
are typically especially vivid, emotional and
not, indeed, be of the same nature as theirs.”
old. Therefore, several critical aspects of these
Such experiences are not limited to the realm
studies do not address the core of Proustian
of artistic license. A survey by the National
phenomena. Firstly, the age of experiences is
Geographic (Gilbert & Wysocki, 1987) gave
limited to time between study and test
readers a set of six odours on scratch-and-sniff
(typically only a few minutes). Secondly,
cards. From a sample of 26 200 (taken
investigators predetermine the target event,
randomly from >1.5 million responses), ~55%
instead of allowing a subject to freely select an
of respondents in their 20s reported at least
autobiographical episode that comes to mind
one emotional memory cued by one of the six
in response to an odour. Lastly, the subject is
odours.
not always made aware of the presence of the
odour during the retrieval phase. Most
Studying odour-cued memories examples of Proustian recall involve an
awareness of the odour.
For empirical substantiation, other paradigms
Many of the approaches used could be
that more closely match the Proustian situation
described as experiments in odour-cued
must be applied. In the standard paired-
context-dependent memory rather than
associate learning paradigm, odours and other
autobiographical memory. Context-dependent
stimuli are associated with target information.
memory is the effect whereby retrieving
Later, the efficacy of each cue type is
information in the same environment in which
compared in a cued recall test. Herz and
Figure 1 Simplified schematic illustrating neuroanatomical and systems-level interactions during the generation of odour-
evoked emotional memories.
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
Cupchik (1995) asked subjects to associate analysis of the pleasantness of memories
odours or odour labels with emotional yielded mixed and inconclusive results.
paintings. Later, they were presented with the Herz and Cupchik (1992) aimed to formulate a
odours or words and asked to describe the characterization of odour-evoked memories
painting with which each had been associated. rather than compare the efficacy of different
A higher degree of emotional tone was found cue types. Their results characterize odour-
in the recollections of paintings cued by evoked memories as highly emotional, vivid,
odours than those cued by words. Here, specific, rare, and relatively old but leave
subjects are actively using odour as an aid to unanswered the question as to whether odour-
retrieval of a past experience. However, evoked memories are, for example, more
autobiographical memories are usually formed emotional. Herz and Schooler (2002) have
through a passive encoding process: separate now answered this question. The emotional
experiential aspects of an incident become and evocative qualities of autobiographical
associated without any purposeful effort on memories cued by odours and visual cues were
the part of the subject. Therefore, there are compared using a new repeated measures
likely to be key differences in the underlying paradigm. Results demonstrated that memories
processes recruited during encoding (and the evoked in the context of odours were
nature of the resulting mnemonic significantly more emotional than those
representation) in the above study as compared recalled in the context of the same cue
with naturally acquired autobiographical presented visually and by the verbal label for
memories. the cue. This is the first unambiguous
demonstration that naturalistic memories
evoked by odours are more emotional than
Examination of ‘true’ those evoked by other cues.
Aggleton and Waskett (1999) provided an
autobiographical memories
excellent demonstration of the potency of
olfactory cues. Their experiment focused on a
Owing to problems of quantitative display at a Viking museum (York) in which
measurement and application of close unusual odours were used to enhance the
experimental controls, only a limited number impact of the display. They asked whether re-
of studies exist which have directly examined presenting these Viking odours to individuals
Proustian phenomena. who had visited the museum a number of years
Laird (1935) wrote about a survey of odours ago would allow them to remember more
‘as revivers of memories and provokers of about the display than those presented with
thoughts in 254 living men and women of non-Viking odours or a no-odour control
eminence.’ More than four out of five of those group. Mean memory performance for those
surveyed reported olfactory-cued memory receiving the Viking odour was numerically
experiences that were described as vivid, higher than for those receiving the non-Viking
emotional and old. 76% of women and 46% of odour and the no-odour control group.
men reported that odour-cued memories were However, due to the expected large variation
amongst their most vivid with only 7% of in performance, the difference failed to reach
women and 14% of men describing these statistical significance.
memories as hedonically neutral. This type of study design is getting closer to a
Contemporary studies employ more scientific fully ecologically valid approach to Proustian
and empirical approaches, for example the phenomena. Naturally occurring experiences
single-cue comparison method. Data is are used and the association between
collected on memories that are retrieved in experience and ambient odour is passively
response to odours and are directly compared formed. Several clear-cut results strongly
with those retrieved in response to stimuli support the proposition that odours are
from other sensory modalities. Rubin et al., effective reminders of autobiographical
(1984) found that odour-cued memories were experience.
thought of and spoken of with less frequency The apparent emotional, vivid and deep-rooted
than those cued by other stimuli. Contrary to nature of odour-evoked memories can be
Proustian expectations, no differences were explained by two hypotheses. (1) The
found between different cue types, and ‘differential cue affordance value’ hypothesis
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
suggests that different sensory modalities vary to be mediated by the action of the amygdala
in terms of cue affordance value (or the (Cahill et al., 1995, 1996).
efficiency with which they access Now equipped with a behavioural background
autobiographical event details). Olfactory cues to odour-cued emotional memory, we move on
are assumed to have a higher cue affordance to analyse the subject from a neurobiological
than other cue types. (2) The ‘differential perspective.
encoding bias’ hypothesis proposes that cues
do not differ in terms of their efficacy in
retrieving event details, but in terms of the The amygdala and olfactory
types of event with which they become
processing
associated and consolidated in memory.
Olfactory details may become associated only,
or preferentially, with more complex The amygdala refers to a highly differentiated
(emotional, personal, and unusual) region near the temporal pole of the
autobiographical episodes. Consequently, the mammalian cerebral hemisphere. Burdach
presentation of an olfactory cue will bias discovered and named the amygdala nucleus in
retrieval towards more complex episodes. the early nineteenth century (Burdach, 1819-
The encoding specificity principle (Tulving & 1822). Subsequent microscopic examination
Thomson, 1973) proposes that consolidated began to reveal a progressively complex
memory traces include not only target structural differentiation. The extent of the
information, but also salient contextual amygdala’s outer border, and the number and
features in the immediate environment in classification of its subdivisions, remains
which the episode was experienced. Later controversial today. Swanson and Petrovich
presentation of such contextual features (1998) suggest that the amygdala is neither a
(which include odours) can therefore act as structural nor a functional unit. One part is a
retrieval cues for related details that comprise specialized ventromedial expanse of the
the original episode. If this alone were the striatum, a second part is caudal olfactory
case, then all cue types would function in the cortex (nucleus of the lateral olfactory tract,
same manner and with equal effectiveness. cortical nucleus, and postpiriform and
However, assuming the superior ability of piriform-amygdalar areas) and a third part is a
olfactory cues to evoke emotional ventromedial extension of the claustrum.
autobiographical memories, encoding Cell groups within the amygdala are derived
specificity alone does not provide a from different regions: their connectivity and
satisfactory explanation. Here lies the the distribution of neurotransmitters within
challenge. each grouping form the basis of
Olfaction is mediated by a number of differentiation. Consequently, different groups
anatomical structures that are also heavily can be distinguished functionally into four
implicated in emotion and memory. The obvious systems – accessory olfactory, main
olfactory bulb, for example, projects to olfactory, autonomic and frontotemporal
structures including the amygdala, cortical.
hippocampus, and thalamus. These structures The amygdala receives highly processed
have been shown to be involved in memory sensory data and is in a position to integrate
function and the modulation of emotions. this data both within and across modalities.
Proustian effects may also be linked to the The only sensory modality with direct access
ability of odours to elicit affective reactions. to the amygdala is smell (see figure two). All
Hinton and Healey (1993) compared reactions other modalities connect with the amygdala
to stimuli presented in visual, lexical, and via a series of sensory association cortices.
olfactory modalities. Odours elicited by far the The amygdala can also influence autonomic,
most affective reactions. hormonal and motor function via its
Some (Herz, 1997; Aggleton & Waskett, subcortical connections.
1999) have ascribed the efficacy of odours in
memory retrieval, at least partially, to the
acknowledged connection between emotional
arousal and the information associated with
such affective reactions. This link is expected
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
Amygdala involvement in non- tests used to date must lack the requisite
specificity to pick up amygdala-dependent
pheromonal forms of olfactory
aspects of olfactory function.
related behaviour
Electrophysiological studies: In one study 229
neurons were recorded in the lateral and
Lesion studies involving rodents: Combined basolateral nuclei of rats during olfactory
electrolytic lesions of the posterior lateral discrimination learning (Schoenbaum et al.,
olfactory tract and anterior amygdala in rats 1999). Of these, 60 were found to show
showed no effect on the retention or learning selective responses depending on whether a
of new olfactory discrimination tasks (Slotnik, particular olfactory cue was paired with a
1985). Bilateral amygdala lesions showed no pleasant or an unpleasant outcome. Selectivity
effect on learning of multiple odour often occurred early in learning and many
discriminations (Eichenbaum et al., 1986). units could be reversed, indicating that the
Failure of amygdala lesions to result in amygdala neurons encoded the motivational
impaired performance seems surprising given significance of the cues.
Main olfactory bulb
Other main olfactory system
Accessory olfactory system
Medial prefrontal,
COApl TR PAA COAa agranular insular
perirhinal,
FS hippocampal cortex
ACB
PA BLAp BMAp BMAa
CEA
MEA
SI
(r) (r,i) (r) (d,r)
MDm
LHAcl THpg
PB
Figure 2 Major neural inputs and outputs of the main olfactory system as they relate to the rat amygdala (adapted from
Swanson and Petrovich, 1998). Abbreviations: ACB, nucleus accumbens; BLAp, basolateral nucleus amygdala, posterior
part; BMAa,p, basomedial nucleus amygdala, anterior, posterior parts; CEA, central nucleus amygdala; COAa,pl, cortical
nucleus amygdala, anterior part, posterior part, lateral zone; d, medial hypothalamic defensive behavior system; FS, fundus
of the striatum; i, medial hypothalamic ingestive behaviour system; LHAcl, lateral hypothalamic area, caudolateral part;
MDm, mediodorsal nucleus thalamus, medial part; MEA, medial nucleus amygdala; PA, posterior nucleus amygdala; PAA,
piriform-amygdala area; r, medial hypothalamic reproductive behavior system; SI, substantia innominata; THpg, thalamus,
perigeniculate region (includes medial geniculate complex, posterior limiting nucleus, and parvicellular subparafascicular
nucleus); TR, postpiriform transition area.
the extensive connections of the olfactory Immediate Early Gene (IEG) studies: c-fos
system illustrated previously and the finding imaging data supplements evidence for
that approximately 40% of rat amygdala involvement of the basolateral amygdala
neurons respond to olfactory stimuli (Cain & (BLA) as olfactory learning progresses (Hess
Bindra, 1972). The lack of obvious deficits et al., 1997). Different stages of olfactory
appears to be caused, in part, by the existence discrimination learning by rats were compared.
of parallel olfactory inputs to other regions Initial exploration (no specific odours) led to
that are able to counterbalance the destruction raised medial (medial, anterior, and posterior
of amygdala olfactory routes. The olfactory cortical nuclei) and, to a lesser extent, raised
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
basolateral activation. In contrast, central matching are severely impaired, although
nucleus levels stayed low. Following initiation H.M. can detect weak odours and appreciate
of cued olfactory discrimination training there odour intensity (Eichenbaum et al., 1983). In
was high basolateral but relatively low medial both instances, it is possible that the
activation. In over-trained rats, the differences combination of amygdala and entorhinal
in these amygdala regions were reduced. pathology is critical for identification. These
Balance of amygdala activation shifts in a and related case studies suggest that the
dynamic manner that depends on the demands involvement of the human amygdala in
of the task and upon the stages of learning. olfaction is often not made apparent by lesion
Lesion studies involving primates: The studies. Presumably, this reflects the presence
contribution of the primate amygdala, and in of multiple olfactory projection routes (i.e.
particular the human amygdala, to olfactory parallel routes to orbitofrontal, thalamic, and
processing remains poorly understood. Group other temporal regions), just as is seen in
studies have shown that comprehensive, rodents. To further investigate involvement of
unilateral temporal lobectomies can confuse the human amygdala in olfaction, the use of
olfactory identification (Rausch & functional imaging studies and the
Serafetinides, 1975) and the recall of odours development of more selective tests is
(Rausch et al., 1977). However, these necessary.
experiments provide limited direct evidence Functional imaging studies: Zald and Pardo
for the specific involvement of the amygdala. (1997) reported that exposure to a highly
As a rule, studies recording the effects of aversive odourant produces a strong increase
surgery to remove just the amygdala or parts in cerebral blood flow as measured by positron
of the amygdala have described no disruption emission tomography (PET) in both
to the sense of smell (Scoville et al., 1953; amygdalae as well as in the left orbitofrontal
Narabayashi, 1977), though these studies lack cortex (see figure three).
proper test details. A more systematic study Interestingly, activity within the left amygdala
has investigated odour discrimination in three was associated significantly with subjective
groups of unilateral temporal lobe resection ratings of aversiveness, suggesting that
cases where the site of surgery was classed as amygdala activity is related to the perceived
either (1) primarily neocortical, (2) hedonic value of the stimulus. Tests conducted
amygdalohippocampectomy, or (3) anterior with pleasant smells (fruits, spices, and florals)
temporal lobe with impingement on the only produced a non-significant increase in
amygdala or hippocampus (Jones-Gotman et blood flow in the right anterior
al., 1997). All three groups were impaired, amygdala/periamygdaloid area. This study
although no clear correlations could be drawn suggests that amygdala activation occurs when
between the presence of an impairment and olfaction engages strong negative emotional
amygdala damage. reactions.
Human electrophysiological studies: These Aversive odours can affect amygdala
have confirmed the responsiveness of activation via associative processes, as has
amygdala neurons to odours (Andy & Jurko, been shown by Schneider et al., (1999) in a
1975) and shown that amygdala activity can functional magnetic resonance imaging (fMRI)
lead to olfactory auras (Andy et al., 1975). A study. The presentation of a neutral face that
small body of single-case clinical studies also had previously been paired with an aversive
implicates the amygdala in aspects of olfactory odour led to a decrease in amygdala activity in
processing. One case involves a woman who a group of 12 normal subjects, whereas
had received an extensive, bilateral increased amygdala activation was found in a
amygdalotomy that had involved the group of 12 subjects suffering from social
entorhinal cortex in one hemisphere (Andy et phobia.
al., 1975). Although her detection of odours Activation of the amygdala by aversive smells
was largely unchanged, her identification of and their associations may suggest other ways
odours was impaired. Similarities are seen in in which smells can influence memory.
patient H.M., in whom the amygdala, Aggleton’s real world study (Aggleton &
entorhinal cortex, and hippocampus were Saunders, 2000) examined the long-term
removed bilaterally (Corkin et al., 1997). retention of specific smells by people who had
Odour discrimination, identification, and previously visited a Viking museum. People
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
who had visited the museum many years al., 1995). A substantial body of evidence
previously almost always recalled that there suggests, however, that lasting
had been smells. However, only those that explicit/declarative memory for emotionally
were independently rated as being most averse arousing information is stored in other brain
were remembered consistently. It was also regions and that the activation of the amygdala
found that the same cues acted as contextual modulates the storage of memory in those
cues to help subjects to recall details of the brain regions (e.g. Cahill & McGaugh 1998;
museum, despite having not visited it for many Weinberger, 1995).
years (Aggleton & Waskett, 1999).

Figure 3 (above and right) Cerebral activation during
aversive olfaction. Changes in regional cerebral blood flow
(rCBF) are rendered in colour with white indicating the
greatest magnitude (Z score > 5) of activation. The relative
positions of coronal sections (A, B, and C) through the
frontal (a and b) and temporal (c) lobes are shown
schematically to the right of this text. Maximal areas of
rCBF change are displayed superimposed on a standard T-
1-weighted magnetic resonance image. The rCBF maxima
map to the amygdala bilaterally and the left posterior
lateral orbitofrontal cortex. The right side of this figure
shows the left side of the brain. Abbreviations: VCA,
vertical line through anterior commissure; CACP,
intercommissural line (reproduced with kind permission
from Zald & Pardo, 1997, © 1997 by PNAS).
Intriguingly, this suggests that the amygdala Hormonal modulation of memory
enhances learning when a subject is
storage
emotionally aroused, and that arousing smells
might also be effective contextual cues
Gerard (1961) proposed “…epinephrine…is
through a similar process.
released in vivid emotional experiences, such
This pathway is a strong candidate for having
that an intense adventure should be highly
a role in the neural underpinnings of Proustian
memorable.” He was the first to advance this
phenomena. An in depth look at arousal-
hypothesis. Gold and van Buskirk (1975)
dependent memory modulation by the
found that systematic post-training injections
amygdala follows.
of adrenaline administered to rats enhances
long-term retention of inhibitory avoidance.
These findings have been confirmed by
The amygdala and modulation of
experiments using many different types of
memory storage: animal studies learning tasks, all in support of the hypothesis
that endogenously released adrenaline
There is now general agreement that the modulates memory storage (e.g. Liang, 2000;
amygdala is involved in the learning of Williams, 2000). As with adrenaline, post-
emotionally arousing information (Cahill et training administration of moderate doses of
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
glucocorticoids also induce dose- and time- clenbuterol into the amygdala produces dose-
dependent modulation of memory storage dependent enhancement of memory storage
(Roozendaal & McGaugh, 1996a). These (Hatfield & McGaugh, 1999). This strongly
findings strongly suggest that glucocorticoids supports the hypothesis that the amygdala
released by strongly emotionally arousing plays a critical role in memory consolidation.
experiences enhance the consolidation of the In the periphery, β-adrenoceptors are located
memory for those experiences (Roozendaal, on vagal afferents that project to the nucleus of
GABAergic agonists and
antagonists
α- and β-adrenoceptor agonists and
antagonists
NTS GABA β cAMP Projections to
STRIA
GR NA other brain
OP α GR TERMINALIS? structures
1
Opioid agonists and
antagonists
BASOLATERAL AMYGDALA
vagus
nerve
BLOOD-BRAIN BARRIER
Peripherally acting β-
adrenergic
Corticosterone
antagonists
Adrenaline
Figure 4 Schematic summarizing the interactions of neuromodulatory influences in the basolateral amygdala on
memory storage. Abbreviations: α1 = α1-adrenoceptor; β1 = β1-adrenoceptor; cAMP = cyclic 3’5’ adenosine
monophosphate; GR = glucocorticoid receptor; NA = noradrenaline; NTS = nucleus of the solitary tract; OP =
opioids (adapted from McGaugh et al., 2000).
2000). the solitary tract (NTS) (Schreurs et al., 1986).
Amygdala mediation of Projections from the nucleus of the solitary
tract are known to release NA within the
neuromodulatory influences on
amygdala (Ricardo & Koh, 1978). Adrenaline
memory storage
does not readily pass through the blood brain
barrier. It seems likely that adrenaline effects
Noradrenaline (NA) release in the brain may memory storage by activation of peripheral β-
regulate memory consolidation (Kety, 1972).
adrenoceptors on vagal afferents. In support of
Gold and van Buskirk’s finding (1978) that
this theory, inactivation of the NTS with
adrenaline and footshock induce the release of
lidocaine blocks adrenaline effects on memory
NA in the forebrain of rats has provided
storage (Williams & McGaugh, 1993).
evidence consistent with Gerard and Kety’s
Recent findings also indicate that the
suggestions. Contemporary studies now
amygdala is involved in mediating
indicate that NA release within the amygdala
glucocorticoid influences on memory storage
is critically important in enabling
(Roozendaal, 2000). Lesions of the stria
neuromodulatory influences on memory
terminalis or of the amygdala block the
storage. Post-training intra-amygdala infusions
memory enhancing effects of post-training
of β-adrenoceptor antagonist propranolol systematic injections of synthetic
block adrenaline effects on memory storage glucocorticoid dexamethasone (Roozendaal &
(Liang et al., 1986) and post-training infusions McGaugh, 1996a,b). These glucocorticoid
of NA or the β-adrenoceptor agonist
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
effects involve noradrenergic activation of the system-activating) learning situations should
amygdala. engage amygdala participation in memory
Noradrenergic activation in the amygdala is formation, independently of whether the
also critical for other neuromodulatory particular emotions involved are positive or
influences on memory storage. Post-training negative. Hence, an emotionally toned
systemic injections of opioid peptides and olfactory stimulus is ‘predisposed’ to
opiates generally impair memory storage and preferential consolidation by the amygdala and
opiate antagonists enhance memory storage therefore improved storage.
(Izquierdo & Diaz, 1983; McGaugh, 1989; Although there is strong evidence in humans
McGaugh et al., 1993). GABAergic for amygdala participation in LTM storage for
antagonists and agonists administered post- emotionally arousing events, evidence for its
training enhance and impair retention, participation in memory retrieval, or for its
respectively (Brioni & McGaugh, 1988; production of emotion, per se, is currently less
Brioni et al., 1989). Lesions of the amygdala clear.
or stria terminalis block opioid peptidergic and
GABAergic influences on memory storage
(McGaugh et al., 1986). Also, and perhaps Emotionally influenced LTM: the
most importantly, intra-amygdala infusions of
role of the adrenergic system
the β-adrenoceptor antagonist propranolol
block opioid peptidergic and GABAergic
Propranolol (a synthetic β-adrenergic receptor
control of memory storage (McGaugh et al.,
blocking agent) attenuates the enhanced LTM
1988). These neuromodulatory interactions
associated with emotionally arousing
within the amygdala are summarized
information without affecting memory for
schematically in figure four.
more emotionally neutral information (Cahill
Studies that have not involved inhibitory
et al., 1994). Hence, activation of the
avoidance training have yielded similar results
adrenergic system in humans is required for
to those summarized above. Hence, there is
the enhancing effect of arousal (either
evidence that amygdala influences on memory
emotionally or physically induced) on
storage are not restricted to the learning of
memory. The action of propranolol is
aversive information.
mediated centrally (van Stegeren et al., 1998).
Adrenergic stimulation can act in a retrograde
manner to enhance memory in humans
Modulation of long-term memory
(Soetens et al., 1995).
(LTM) in humans: adrenergic
activation and the amygdala
Emotionally influenced LTM: the
I begin here with the assumption that the brain
role of the amygdala
must possess a means to ‘weight’ or moderate
the actual information storage mechanisms in
Scoville and Millner examined memory in ten
approximate proportion to the importance of
patients, including some with relatively
the information being stored (Gold &
selective amygdala damage and concluded
McGaugh, 1975). Without modulatory
“removal of the amygdala bilaterally does not
mechanisms, the brain would be unable to
appear to cause memory impairment”
distinguish the important from the trivial in
(Scoville & Millner, 1957). However, their
LTM. There appear to be, at minimum, two
tests involved short-term memory, and/or
interacting neurobiological components
memory for non-emotionally arousing
mediating the modulatory actions of emotional
material. These findings fit well with the
arousal on memory: the adrenergic system and
‘memory modulation’ view of amygdala
the amygdala (especially BLA). An interaction
function, which predicts that amygdala activity
between endogenous stress hormones activated
should not be critical for STM, or for LTM of
by emotional experiences and the amygdala
non-emotionally arousing events or material.
appears critical. According to animal studies,
A rare patient with damage confined almost
we can hypothesize that sufficiently
exclusively to her amygdala reported deficits
emotionally arousing (i.e. sympathetic nervous
Pages 18-37
Olfaction, Emotion and the Amygdala: Hughes
in memory for emotional material (Babinsky et observed either pleasant or unpleasant
al., 1993). Enhanced LTM associated with an arousing images. Memory for the images one
emotionally arousing story is impaired in month later correlated very highly with
patients with selective (Cahill et al., 1995) or amygdala activity while viewing either the
nearly selective (Adolphs et al., 1997) pleasant or unpleasant images. Amygdala
amygdala damage. Amnesic patients with activity whilst viewing emotionally neutral
undamaged amygdalae show reasonably intact images (including images of novel, but not
enrichment of memory for emotional data emotionally arousing objects) did not correlate
despite their overall impaired memory with recall. The arousing dimension of the
performance (Hamann et al., 1997). The pictures was essential to amygdala
emotional reactions of patients with amygdala engagement in memory, as indicated in
damage to emotionally provocative stimuli are another recent fMRI study (LaBar et al.,
not significantly different from those of 1998).
controls even though LTM for emotional Although activity of both the left and right
material is impaired (Cahill et al., 1996). amygdala correlated with recall of arousing
Electrodermal responses to emotionally material, the degree appeared greater on the
stressful events are intact in humans with right for negative slides (Hamann et al., 1999).
amygdala damage (Adolphs et al., 1997). On Cahill et al. (1996), who used emotionally
the basis of these and PET studies described, negative material and found a unilateral effect,
Cahill et al., (1996) propose that amygdala also suggest that differences in left versus right
activity in humans “may be more important amygdala participation in memory for
for the translation of an emotional reaction emotionally arousing events may relate to their
into heightened recall than it is for the pleasant/unpleasant nature. Left amygdala
generation of an emotional reaction per se.” activation appears to be produced consistently
Brain imaging studies: In one study, subjects by any of a variety of aversive stimuli used in
received two PET scans: one while viewing a human brain imaging studies (Cahill &
series of relatively emotionally arousing films, McGaugh, 1998; Zald & Pardo, 1997, 2000 for
another while viewing a series of relatively olfactory examples). No compelling
emotionally neutral films (Cahill et al., 1996). explanation exists at present for such
Memory for the films was tested in a surprise asymmetries in amygdala function.
free recall test three weeks later. Activity in Canli and colleagues (1999) reported that
the right amygdala correlated very highly amygdala activity in response to emotionally
(r=0.93) with the number of emotional films negative pictures correlated with long-term (2-
recalled three weeks later, but not with recall 14 months) retention of pictures. In an
of the neutral films (see figure five). individual fMRI study conducted in
Figure 5 (A) Amygdala activity while watching a series of emotionally arousing films correlated very highly with long-
term (three-week) recall of the films. (B) Amygdala activity in the same subjects while viewing a series of relatively
emotionally neutral films did not correlate significantly with recall (reproduced with kind permission from Cahill et al.,
1996, © 1996 by PNAS).
Hamann and colleagues (1999) used PET to collaboration with the above, it was found that
study cerebral blood flow while subjects increasing the degree of emotional arousal