ll

Review
Mechanisms of memory under stress
€ls,4,5 and Benno Roozendaal2,3
Lars Schwabe,1,* Erno J. Hermans,2,3 Marian Joe
1Department €t Hamburg, Hamburg, Germany
of Cognitive Psychology, Universita
2Donders Institute for Brain, Cognition and Behaviour, Radboud University, Nijmegen, the Netherlands
3Department of Cognitive Neuroscience, Radboud University Medical Center, Nijmegen, the Netherlands
4University Medical Center Groningen, University of Groningen, Groningen, the Netherlands
5UMC Utrecht Brain Center, Utrecht University, Utrecht, the Netherlands

*Correspondence: lars.schwabe@uni-hamburg.de
https://doi.org/10.1016/j.neuron.2022.02.020

SUMMARY

It is well established that stress has a major impact on memory, driven by the concerted action of various
stress mediators on the brain. Recent years, however, have seen considerable advances in our understand-
ing of the cellular, neural network, and cognitive mechanisms through which stress alters memory. These
novel insights highlight the intricate interplay of multiple stress mediators, including—beyond corticoste-
roids, catecholamines, and peptides—for instance, endocannabinoids, which results in time-dependent
shifts in large-scale neural networks. Such stress-induced network shifts enable highly specific memories
of the stressful experience in the long run at the cost of transient impairments in mnemonic flexibility during
and shortly after a stressful event. Based on these recent discoveries, we provide a new integrative frame-
work that links the cellular, systems, and cognitive mechanisms underlying acute stress effects on memory
processes and points to potential targets for treating aberrant memory in stress-related mental disorders.

INTRODUCTION shifts—recently uncovered changes in memory dynamics

and flexibility.
Stress effects on memory are driven by the numerous neuro-
Stressful events are arguably the most important ones to
transmitters, hormones, and peptides that are released in
remember. An animal has to be able to tell immediately
response to stressful events and act directly, or indirectly via
whether sounds, places, scents, and other animals are
brainstem circuits, on medial-temporal and prefrontal areas
dangerous. An animal that has to mentally rehearse the
crucial for memory (Joe €ls and Baram, 2009; Figure 1). Altogether,
sounds and smells of, say, a bushfire, is not likely to sur-
these stress mediators synergistically promote coping with an
vive.—Bruce S. McEwen (2002, Page 108), 1938–2020
ongoing stressor by supporting an initial ‘‘fight-or-flight’’
In medieval times, communities threw young children in the response that allows the individual to respond appropriately to
river when they wanted them to remember important events. the situation at hand, followed by a later phase geared to ratio-
They believed that throwing a child in the water after witness- nalize and store the information linked to its context. Thus,
ing historic proceedings would leave a lifelong memory for the stress-induced changes in memory processes are an integral
events in the child (McGaugh, 2003). Although this cruel tradi- part of the behavioral adaptation to stressors. These changes
tion stopped—fortunately—centuries ago, modern research initially lead to prioritized attentional and appraisal processing
confirms that stressful or arousing experiences may indeed of emotionally salient events, increase the reliance on well-es-
boost memory for surrounding events (McGaugh, 2015). tablished habits and routines, reduce distraction by stressor-
Research over the past decades, however, painted a much irrelevant information, and—in the aftermath of stress—promote
more nuanced picture of how stressful events shape memory, the storage of information most relevant of the stressful
showing that stress enhances some memory processes but encounter to facilitate coping with similar future events (Diamond
impairs others, and that different stress response patterns et al., 2007; Joe €ls et al., 2006; Vogel et al., 2016). While being
associated, e.g., with different types of stressors, may affect generally highly adaptive, overly strong or aberrant stress effects
what information is being encoded and how it is stored. More- on cognitive processing, particularly on memory formation, can
over, research in rodents and humans provided exciting in- become maladaptive and contribute to stress-related mental
sights into the brain mechanisms underlying the impact of disorders such as posttraumatic stress disorder (PTSD) or anxi-
stress on memory. In this review, we will discuss recent dis- ety disorders (de Quervain et al., 2017; Pitman et al., 2012).
coveries in the field that have transformed our thinking of These clinical implications of stress effects on memory may
how stress affects memory. From these findings, we will have contributed to the enormous amount of research in this
derive a new integrative framework of how acute exposure area over the past decades, with different and sometimes para-
to a stressful event initiates—through the orchestrated action doxical views (Figure 2). Research on the stress-memory link
of multiple stress mediators and specific neural network was stimulated half a century ago by the seminal discovery

1450 Neuron 110, May 4, 2022 ª 2022 Elsevier Inc.

 ll
Review

Figure 1. Multiple stress response systems

Within (milli)seconds after stressor onset, the release of monoamines, including dopamine (DA), noradrenaline (NA), and serotonin (ST), is increased in specific
neuronal populations. Neurons in hypothalamic nuclei further rapidly activate the sympathetic nervous system, which triggers the release of adrenaline and NA
from the adrenal medulla. In parallel, the hypothalamus stimulates the slower hypothalamus-pituitary-adrenal (HPA) axis, a hormonal cascade that includes the
release of corticotropin-releasing factor (CRF), vasopressin and adrenocorticotropic hormone (ACTH) and leads within minutes to the secretion of corticosteroids
(e.g., cortisol in humans, corticosterone in rodents) from the adrenal cortex. These multiple stress mediators are thus released in waves, reaching the brain at
different time points (top). Each of the multiple stress mediators has its specific temporal profile of action on the brain (as indicated by the arrows). The temporal
windows of action may overlap, thus enabling synergistic actions between stress mediators. For instance, while corticosteroids were traditionally thought to act
via intracellular mineralocorticoid (MR) and glucocorticoid receptors (GRs) leading to slow, genomic actions, it is by now established that corticosteroids exert
their actions also via near-membrane MR and GR, which enable rapid, non-genomic actions allowing interactions with the fast-acting noradrenergic system.

that corticosteroids can enter the brain and that their receptors dependent effects on memory (Akirav et al., 2004; Sandi et al.,
(i.e., glucocorticoid receptors [GRs] and mineralocorticoid re- 1997), partially due to the balance or imbalance of MR- and
ceptors [MRs]) are expressed at particularly high density in the GR-mediated actions (de Kloet et al., 1999). The subsequent dis-
hippocampus (McEwen et al., 1968; Reul and de Kloet, 1985), covery that corticosteroids act not only via nuclear receptors
a key region for memory (Squire, 1992). Based on subsequent mediating slow genomic actions but also via near-membrane re-
findings showing that stress or corticosteroids can block hippo- ceptors allowing non-genomic actions showed that corticoste-
campal synaptic plasticity (Diamond and Rose, 1994; Pavlides et roid effects can unfold much more rapidly than previously thought
al., 1995) and impair hippocampus-dependent spatial or declar- (Dallman, 2005; Di et al., 2003; Karst et al., 2005). This non-
ative memory (Diamond and Rose, 1994; Lupien et al., 1997; genomic mode of action enables corticosteroid interactions
Newcomer et al., 1994), the view at the end of the past century with the rapidly acting noradrenergic system, which represents
held by some was that corticosteroids disrupt memory (whereas a key mechanism through which stress enhances the consolida-
catecholamines may enhance amygdala-dependent memory; tion (Cahill et al., 2003; Roozendaal et al., 2006), but impairs the
McEwen and Sapolsky, 1995), although this was not unequivocal retrieval of memory (Buchanan et al., 2006; de Quervain et al.,
(Oitzl et al., 1997; Roozendaal and McGaugh, 1996). 1998, 2000; Roozendaal et al., 2004). Together, these findings
This idea of a global stress- or corticosteroid-induced memory suggested that stress enhances memory for material encoded
deficit changed decisively when corticosteroid actions via the within the context and around the time of the stressor, when
MR and GR as well as their interactions with the noradrenergic (non-genomic) corticosteroid, (nor)adrenergic and potentially
system were better understood, pointing to dose- and time- neuropeptide activity are synchronized, but impairs memory for
dependent effects of stress and stress hormones. In particular, information that occurs out of context, when the activations of
it was shown that GRs and MRs have distinct functions in mem- major stress response systems are desynchronized, i.e., do not
ory (Oitzl and de Kloet, 1992) and that corticosteroids exert dose- overlap (Diamond et al., 2007; Joe €ls et al., 2006, 2011).

Neuron 110, May 4, 2022 1451

 ll
Review

Figure 2. Historic progression of research on stress effects on memory

Overview of significant advances in our understanding of how stress and stress mediators affect memory and its neural underpinnings (bottom) and related
changes in the predominant view of how stress shapes memory (top).
GC, glucocorticoids; LTP, long-term potentiation; GR, glucocorticoid receptor; MR, mineralocorticoid receptor; ECBs, endocannabinoids.

Building on these cornerstones, recent years have seen memory and executive functions, or in cognition across the
important advances that provide new insights into the mecha- lifespan—have been covered in excellent previous reviews (An-
nisms involved in stress effects on memory—especially in the dreano and Cahill, 2009; Arnsten, 2009; Conrad, 2010; Lupien
human brain—and, at the same time, show that these effects et al., 2009; Shields et al., 2016).
are much richer and more complex than previously thought.
For instance, it is becoming increasingly clear that the impact CELLULAR MECHANISMS UNDERLYING THE IMPACT
of stress on memory depends, as outlined below, critically on OF STRESS ON MEMORY
the history or state of the individual, e.g., a naive individual
versus one that has recently experienced an acute stressor, Understanding the mechanisms by which acute stress affects
and that this impact relies on large-scale network interactions memory formation starts with the notion that a physiological
rather than effects on isolated brain areas. Moreover, stress ef- stress response can be deconstructed into different waves of
fects on memory may only be understood when taking the differ- stress mediators, sequentially reaching and affecting brain cells
ential contributions of multiple anatomically and functionally (Figure 1). Stress mediators include catecholamines, such as
distinct memory systems into account and that stress will influ- adrenaline and noradrenaline (NA); neuropeptides, such as corti-
ence memory over a wide range of time, from the initial memory cotropin-releasing factor (CRF); and corticosteroid hormones,
encoding phase, through consolidation to storage, even months primarily corticosterone in rodents and cortisol in humans (Joe €ls
after the stressful event. and Baram, 2009). Exactly which cells or brain regions are
In this review, we will give an overview of the current state of affected depends on (1) whether or not particular brain areas
thinking of how acute stress affects memory, with an emphasis are reached by stress mediators, (2) the local expression of re-
on recent insights, at multiple levels of integration—from cells, ceptors, (3) the type and severity of the stressor, and (4) the na-
through microcircuits and local systems, to whole-brain and €ls et al., 2012). How neuronal activity
ture of the learning task (Joe
cognitive consequences—in rodents and humans. These recent is altered also depends on signaling pathways downstream of
advances will then be integrated with previously established these receptors and the cellular context. Generally, catechol-
mechanisms to provide a new integrative framework that will amines and neuropeptides act within minutes through mem-
link the cellular, neural network, and cognitive levels of the brane receptors while corticosteroids exert delayed effects,
impact of stress on memory processes, from memory consolida- binding to intracellularly located receptors that serve as tran-
tion and retrieval to memory flexibility and dynamics. We primar- scription factors, although more recently also rapid, non-
ily focus on the effects of acute stress on long-term memory pro- genomic signaling was revealed. Altogether, stress mediators
cesses. Effects of chronic stress on memory—same as other can change cell activity over a wide range of time, from minutes
issues that we lack space to discuss in detail here, such as up to hours and days, partly explaining why even a brief stressor
potential sex differences, stress-induced changes in working can change later phases of memory formation.

1452 Neuron 110, May 4, 2022

 ll
Review

Box 1. Interaction of multiple stress mediators in shaping memory processes

Studies in rodents have provided a beautiful illustration how one can make the step from biochemical signaling at the level of cells
and microcircuits to cognitive processing and behavioral output, serving as a bridge ultimately to human brain networks and cogni-
tive processing. These in vivo studies allow investigation of multiple brain areas and levels of integration.
Animal studies have shown functional interactions between corticosterone and NA on memory. Corticosterone administration to
rats after footshock delivery in an inhibitory avoidance task rapidly augments NA levels within the BLA (McReynolds et al., 2010). In
contrast, attenuation of noradrenergic signaling with b-adrenoceptor antagonists infused into the BLA blocked the memory
enhancement induced by a corticosteroid administered either systemically or directly into a variety of other brain regions such
as the HP or PFC (Barsegyan et al., 2010; Quirarte et al., 1997; Roozendaal et al., 2002, 2006). CRF effects on memory are also
dependent on interactions with both the noradrenergic and corticosteroid systems (Roozendaal et al., 2008). These interactions
of corticosteroids and CRF with the noradrenergic system may provide a direct explanation for the finding that these stress me-
diators selectively enhance memory consolidation of emotionally arousing experiences (Buchanan and Lovallo, 2001; Cahill
et al., 2003; Okuda et al., 2004; Roozendaal et al., 2006).
Several experimental findings suggested that corticosteroid interactions with noradrenergic signaling might have an onset that is
too fast to be mediated via transcriptional regulation in the nucleus and likely involve rapid, non-genomic interactions with the ECB
system. Stressful training or a single injection of corticosterone rapidly elevates ECB levels in corticolimbic regions (Hill et al., 2010;
Morena et al., 2014). Conversely, a CB1 receptor antagonist administered into the BLA blocked the enhancing effect of posttraining
systemic corticosterone on memory consolidation (Campolongo et al., 2009). Further, a CB1 receptor antagonist infused into the
BLA blocked the memory-enhancing effects induced by either a specific GR agonist or the membrane-impermeable ligand
cort:BSA (Atsak et al., 2015), indicating that corticosteroid-ECB interactions on memory presumably involve the activation of a
GR on or near the cell surface. Although the initial studies examining corticosteroid interactions with the ECB system on memory
consolidation have focused on the BLA, subsequent studies have shown highly comparable interactions within the HP, PFC, and
dorsal striatum (Morena et al., 2014; Siller-Pérez et al.,2019). Moreover, several studies have shown that corticosteroid effects on
retrieval impairment also require an interaction with the ECB system (Atsak et al., 2012; Morena et al., 2015). Intriguingly, whereas
corticosteroid-ECB interactions on memory consolidation appear to predominantly involve the ECB ligand anandamide (Morena
et al., 2014), corticosteroid effects on memory retrieval have been shown to depend on 2-arachidonoylglycerol signaling (Morena
et al., 2015).
Subsequent experiments indicated that such corticosteroid-induced recruitment of the ECB system is also critically involved in
regulating the rapid effects of corticosteroids onto the noradrenergic system (Atsak et al., 2015). The ECB system might either
(Continued on next page)

Neuron 110, May 4, 2022 1453

 ll
Review

Box 1. Continued

directly influence noradrenergic activity or, alternatively, alter noradrenergic function indirectly via a modulation of GABAergic or
glutamatergic activity. Within the BLA, CB1 receptors are in particular abundantly expressed on GABAergic interneurons (Katona
et al., 2001) and activation of CB1 receptors has consistently been shown to suppress the release of GABA (Ohno-Shosaku et al.,
2001). Suppressing GABA activity is known to stimulate the release of NA (Hatfield et al., 1999). Together, these findings thus sug-
gest that corticosterone might bind to a GR on the cell surface and rapidly induce the release of ECBs. The released ECBs then bind
to CB1 receptors on GABAergic interneurons and inhibit the release of GABA that can then result in a change in excitation/inhibition
balance and a disinhibition of noradrenergic transmission in BLA neurons (Di et al., 2016).

Although in principle many brain areas are reached by one or Clayton et al., 2020). Thus, stress mediators have a quick and
more stress mediators, most molecular, biochemical, and elec- strong excitatory effect on particularly BLA neurons, while in hip-
trophysiological studies have been confined to a limited set of pocampal and particularly PFC principal cells the signal-to-noise
areas, i.e., subregions of the hippocampus and prefrontal cortex activity is improved in a slow GR-dependent manner involving
(PFC), the amygdalar nuclei and to a lesser extent the nucleus gene transcription (Joe €ls et al., 2018). The timeframe of these
accumbens, ventral tegmental area, and hypothalamus (Bains rapid and slow effects would allow modulation of memory en-
et al., 2015; Joe€ls et al., 2012; Peng et al., 2021). This choice coding and consolidation, respectively, yet our knowledge is still
was guided by the behavioral or endocrine relevance of these limited with respect to the modulation of the neurocognitive pro-
areas and is clearly a limitation in our body of knowledge. Gener- cesses subserved by the regions that have been in the spotlight
ally, NA increases excitatory transmission and synaptic plasticity so far, and at the cellular level these regions have been mostly
through b-adrenoceptors, although a role of a-adrenoceptors is examined in isolation.
also indicated (Arnsten, 2009; Ferry et al., 1999). Similarly, CRF Nearly all in vitro electrophysiological studies so far focused on
through CRF1 receptors mostly enhances limbic excitability glutamatergic transmission (mostly in principal neurons), which
and synaptic plasticity. For instance, acute stress and enhanced seems justified since GR knockout in glutamatergic neurons is
CRF levels in general cause rapid remodeling of CA1 hippocam- important for, e.g., abolition of anxiogenic effects of stress,
pal spines, promote glutamate release, and improve synaptic whereas GR knockout in GABAergic neurons proved to be inef-
plasticity (Vandael et al., 2021). Conversely, CRF2 receptors fective (Hartmann et al., 2017). Yet, several recent studies under-
are involved in the termination of the stress response (Henckens line that this may give an incomplete picture. For instance, after
et al., 2016). Corticosteroids also yield a differentiated picture. In elevated platform stress GABAergic currents as well as the inhi-
the mouse hippocampus (Karst et al., 2005), corticosterone bition-to-excitation ratio were reduced in CA1 hippocampal neu-
quickly but reversibly increases glutamate release probability, rons, which was related to impaired retrieval of spatial memory
through a non-genomic route involving the MR. This is paralleled after acute stress (Shi et al., 2020). As mentioned, indirect corti-
by an MR-dependent increase in GluR2-AMPAR surface diffu- costeroid effects on GABAergic interneurons in the BLA via
sion (Groc et al., 2008). In the basolateral amygdala (BLA) too, retrogradely transported ECBs play a crucial role in the
corticosterone quickly increases glutamate transmission quick boost of local inhibitory transmission (Di et al., 2016; see
through MR but here the effects are long-lasting (Karst et al., Box 1). Recent evidence also supports a role of GR and ECBs
2010). More recently, it has become evident that glutamate in mitochondrial function that could relate to memory formation
signaling of principal cells in the BLA is also boosted via rapid (Hebert-Chatelain et al., 2016) and anxiety (Filiou and Sandi,
GR-dependent activation of endocannabinoids (ECBs)—up- 2019). The relevance of GABAergic transmission and local cir-
stream of noradrenergic signaling—which then retrogradely cuitry for long-term emotional memory formation furthermore
through inhibition of GABAergic cells causes local disinhibition emerged from a recent study showing that de novo translation
(Campolongo et al., 2009; Di et al., 2016; see Box 1). in somatostatin-expressing centrolateral amygdala interneurons
The rapid-onset corticosteroid effects are complemented is necessary for the consolidation of conditioned threat re-
by late, genomic actions via GRs, e.g., on glutamate signaling, sponses, which is distinct from the pathway involved in dimin-
causing increased glutamate responses in CA1 pyramidal ished responses to a safety cue, which depends on translation
neurons (Karst and Joe €ls, 2005) and layer V PFC cells in another set of inhibitory neurons (Shrestha et al., 2020).
(Yuen et al., 2009), while other forms of transmission are gener- Not only principal neurons and interneurons are affected
ally suppressed (Joe €ls et al., 2012). In hippocampal cells, GR by stress mediators; it has become increasingly evident that
activation also enhances synaptic dwell time of diffusing (micro)glial cells might also be implicated in effects of acute stress
GluR2-AMPARs (Groc et al., 2008). Late effects, which on memory. For instance, mice with astrocyte-specific GR dele-
take >1 h to develop, involve altered gene transcription, but tion showed impaired aversive memory expression (Tertil et al.,
to date, the signaling cascades—from receptor to effector 2018). This may involve altered glucose uptake in astrocytes
molecule—remain elusive, despite many efforts to delineate by the glial isoform of serum/glucocorticoid regulated kinase 1
these pathways, focusing e.g., on candidate molecules (Sgk1). Alternatively, this effect might be related to NMDA-depen-
such as cAMP response element-binding protein (CREB; Buur- dent long-term potentiation (LTP) in hippocampal astrocytes
stede et al., 2021) and tissue plasminogen activator (tPA; during task acquisition (Adamsky et al., 2018). Thus, part of the
Bouarab et al., 2021), or investigating the entire genome (see memory-promoting effects of stress or corticosteroids could be

1454 Neuron 110, May 4, 2022

 ll
Review

accomplished through astrocytes, as part of a tripartite synaptic from the amygdala, is necessary and sufficient to induce anxi-
complex (Popoli et al., 2011). ety-like behavior in an open field or elevated zero-maze; a poten-
The approach to investigate one stress mediator or one area at tial cellular underpinning of the observation in humans that a
a time has been helpful to generate a theoretical framework of salience processing network is involved in the initial stages of
how acute stress might alter local network function (Joe €ls stressful learning (see below). Moreover, activation of a dense
et al., 2018). Nevertheless, it is a very reductionistic approach noradrenergic projection from the locus coeruleus to the dentate
and does not do justice to (1) the complexity of the stress gyrus resulted in contextual generalization through b-adren-
response and its multiple mediators, (2) the fact that stress ef- ergic-mediated modulation of hilar interneurons (Seo et al.,
fects depend on the history and state-dependent characteristics 2021). The importance of NA (as one of the stress mediators)
of the animal, including the state induced by the learning task, for fear learning was also revealed using an activity-dependent
and (3) the notion that learning tasks involve integrated networks tagging system (Leal Santos et al., 2021): the b-adrenoceptor
of brain areas that collectively lead to encoding and memory for- antagonist propranolol, which blocks lower-affinity b-adreno-
mation. An illustration of the first issue is the fact that waves of ceptors that are occupied at higher tonic levels of NA observed
stress mediators overlap in time and space (Joe €ls and Baram, directly after acute stress, acutely impaired fear memory traces
2009), and one wave may affect the cellular response to the and altered functional connectivity between the dorsal dentate
next (‘‘metaplasticity’’). This principle was illustrated for BLA gyrus, PFC, and BLA. These studies elegantly support the earlier
neurons, where activation of b1-adrenoceptors suppressed the in vivo observation that the BLA and dentate gyrus are important
electrophysiological response to corticosterone administered hubs in mediating interactive effects of NA and its interaction
20 min later (Karst and Joe €ls, 2016). Consequently, low to mod- with corticosteroids on synaptic plasticity (Vouimba et al., 2007).
erate concentrations of b1-adrenoceptor agonists and cortico-
sterone resulted in curtailed excitatory BLA responses, while LARGE-SCALE NETWORK INTERACTIONS
high concentrations resulted in lengthy activation. UNDERLYING THE IMPACT OF STRESS ON MEMORY
Metaplasticity also comes into play with repeated peaks of
corticosterone, showing that cellular responses to corticoste- Extensive evidence indicates that the different stress mediators
roids may differ depending on whether they take place in a naive enhance memory by acting within many different brain regions.
animal or one that has recently experienced an acute stressor; Notably, however, these brain regions are highly functionally in-
this emphasizes the relevance of the history and state-depen- terconnected (McGaugh, 2000; Roozendaal and McGaugh,
dent characteristics of the animal. Thus, corticosteroid exposure 2011). For instance, previous rodent studies have implicated
of amygdalar cells in recently stressed mice ‘‘decreased’’ gluta- the BLA in orchestrating memory-enhancing effects of these
matergic transmission via GR, as opposed to the MR-dependent stress mediators, not only by modulating neuroplasticity and
increase in glutamatergic transmission seen in naive mice (Karst memory processes elsewhere in the brain (Barsegyan et al.,
et al., 2010). A similar metaplastic switch, now for synaptic plas- 2019; Bonapersona et al., 2022; Chen et al., 2018; Ikegaya
ticity, was seen with respect to auditory fear conditioning in the et al., 1997; Lovitz and Thompson, 2015; McIntyre et al., 2005;
lateral amygdala (Inoue et al., 2018). Also ultradian corticoste- Roozendaal and McGaugh, 2011), but also by enabling direct
rone pulses of variable amplitude, at different phases of the stress hormone effects in other brain regions, and thereby influ-
circadian rhythm (Lightman et al., 2020), can metaplastically encing functional interactions within larger brain networks (Bar-
change spontaneous BLA glutamate transmission, which could segyan et al., 2019). Such observations of widespread
explain why tone-cue fear conditioning is most effective during network-level changes dovetail with observations from func-
the inactive phase of the day (den Boon et al., 2019). tional neuroimaging in humans, which indicate that memory for-
With respect to the third cause of complexity—the existence mation is supported by activity across networks that span the
of interactive networks—it has become clear that many cells entire brain (Ranganath and Ritchey, 2012). This neuroimaging
and brain regions will show changed activity after acute stress work has revealed that certain network configurations are
(Bonapersona et al., 2022), and it is the collective and integrated required to guide attention to salient stimuli and support mne-
response in entire networks that determines the overall rele- monic operations that form initial memory traces, while other
vance for cognitive processing. In vitro experiments are ill-suited network configurations critically support consolidation, transfor-
to study integrated effects of multiple brain areas at a time. To mation, and long-term storage of information. As we will
really appreciate the cellular effects accompanying acute stress describe below, stress-related neuromodulatory actions appear
and their relevance for memory formation, in vivo recordings are to play a critical role in guiding these network interactions and
indispensable. Functional MRI in principle could provide a switches (Figure 3).
whole-brain overview of activity in rodents—thus bridging the One large-scale network identified using human functional
methodology in animals and humans—yet this method is inher- neuroimaging is the ‘‘salience’’ network (SN; Seeley et al.,
ently stressful to rodents and therefore not suitable. Moreover, 2007). This network prominently includes the amygdala, but
this approach does not provide information at the single-cell also encompasses dorsal anterior cingulate/dorsomedial PFC,
level. Simultaneous single-cell recordings across multiple areas anterior insula, temporoparietal junction, thalamus, striatum,
will need to give an answer, and—though sparse—recent and hypothalamus. It is thought to integrate neurocognitive
studies indeed give more insight. For instance, McCall et al. systems required for an optimal response to homeostatic
(2015) demonstrated that increased tonic activity of the locus co- threats at all stages from optimizing sensory intake and initial
eruleus noradrenergic system, depending on CRF projections appraisal to generating appropriate responses (Seeley, 2019),

Neuron 110, May 4, 2022 1455

 ll
Review

A stress-induced shift toward enhanced coupling between

amygdala and striatal regions within the SN was furthermore
shown to be reduced in carriers of a functional deletion variant
of the gene encoding the a2b-adrenoceptor, resulting in
increased NA signaling (Wirz et al., 2017b). Chemogenetic and
optogenetic studies in rodents in recent years have borne out
the causal link between LC-NA activation and anxiogenesis
(Hirschberg et al., 2017), and increased connectivity within (a ro-
dent homolog of) the SN (Zerbi et al., 2019). In addition to norad-
renergic signaling, recent evidence indicates that in the immedi-
ate phase of the stress response, corticosteroid action via MR
may have a synergistic effect with NA (Vogel et al., 2016; Wirz
et al., 2017a). Together, these findings suggest that stress-
induced tonically elevated levels of NA, potentially in synergy
with rapid non-genomic corticosteroid effects via MR, switch
the brain to a stimulus-unselective hypervigilant ‘‘encoding’’ or
‘‘memory formation’’ mode. These rapid effects congrue with
the framework provided by the (earlier) cellular studies.
While unselective encoding of as much information as
possible during a stressful experience may be adaptive, it
Figure 3. Stress-induced shift of large-scale neural networks
Acute stress leads to a rapid reconfiguration of large-scale neural networks, would run into capacity limits, interference, and poor signal-
favoring the salience network over executive control and default-mode net- to-noise ratio when prolonged for too long. A critical question
works. The late phase of the stress response, however, may evoke a reversal is therefore how the brain balances the need of retaining all
of this network shift, now promoting the executive control and default-mode
networks over the salience network. potentially relevant information with the need to avoid exces-
dlPFC, dorsolateral prefrontal cortex; mPFC, medial prefrontal cortex; pPC, sive storage of irrelevant material. The solution appears to lie
posterior parietal cortex; MTL, medial-temporal lobe; pCC, posterior cingulate in a comprehensive reconfiguration of large-scale network
cortex; dACC, dorsal anterior cingulate cortex; AM, amygdala; fI, frontal insula;
activity that initiates robustly and immediately after external de-
MID, midbrain; iT, inferior temporal cortex.
mands (e.g., due to a stressor) subside. This switch to an ‘‘off-
line’’ mode, away from externally and toward internally directed
including memory encoding (Hermans et al., 2014a). Indeed, cognition, involves robust and consistent activation within (ven-
human neuroimaging work has shown that activity of (key re- tro)medial PFC, inferior parietal, posterior cingulate, and
gions within) the SN is associated with subsequent memory retrosplenial regions, which are together referred to as the
retention specifically for emotionally arousing material (Hamann default-mode network (DMN) (Raichle, 2013). Notably, the
et al., 1999; Kim, 2011) and also predicts later involuntary intru- DMN also exhibits strong intrinsic functional connectivity with
sions (Visser et al., 2021). Thus, SN activation results the medial-temporal lobe (MTL), including both hippocampus
in prioritized encoding of stress-relevant over peripheral and amygdala (Buckner et al., 2008). Although its precise func-
information. tion remains debated, there is growing consensus regarding a
SN activation appears to be tightly coupled to noradrenergic role for the DMN in mnemonically related operations such as
signaling. In particular phasic activity of the locus coeruleus- prospection or, more broadly, ‘‘offline’’ associative processing
noradrenaline (LC-NA) system, the main source of NA in the (Bar, 2021). These notions concur with findings in both rodents
brain, is thought to engage SN regions to prompt task-set and humans of persistent experience-specific activity patterns
switches in response to salient stimuli (Corbetta et al., 2008). following learning in hippocampal-cortical circuits (Ji and Wil-
CRF is well known to interact with other stress mediators to regu- son, 2007; Tambini and Davachi, 2019), including medial PFC
late tonic and phasic activity of the LC and its communication (Takehara-Nishiuchi and McNaughton, 2008; Van Kesteren
with key regions of the SN such as the amygdala (Valentino et al., 2010). DMN activation is therefore thought to support
and van Bockstaele, 2005). Further, as argued above, CRF pro- integration of novel information into existing associative ‘‘sche-
jections from the amygdala can trigger a switch toward a tonic mas’’ and thereby facilitate early stages of the gradual shift to-
mode of LC-NA firing, which diminishes phasic firing to discrete ward cortico-cortical dependency of memory that is referred to
stimuli, has an anxiogenic effect, and creates a hypervigilant and as systems consolidation (Gilboa and Moscovitch, 2021).
distractible attentional state (McCall et al., 2015). Functional neu- Similar to the switch to an ‘‘encoding’’ or ‘‘memory formation’’
roimaging work in humans has shown that functional connectiv- mode, the early stage of systems consolidation appears to be
ity within the SN is increased during exposure to highly nega- tightly controlled by stress-related activation of the LC-NA sys-
tively arousing cinematographic material (Hermans et al., tem and its effect on amygdala-centered networks, likely in syn-
2011). Synchronization of activity within this network also fluctu- ergy with corticosteroid actions. Notably, such effects occur
ated dynamically with levels of physiological arousal induced by during time windows in which arousal-related noradrenergic ac-
these film clips (Young et al., 2017). Stress-induced SN connec- tivity remains tonically elevated following stressors. In line with
tivity furthermore diminished following administration of the electrophysiological studies in rodents (Pape and Paré, 2010;
b-adrenoceptor antagonist propranolol (Hermans et al., 2011). Paré, 2003; Popa et al., 2010; Seidenbecher et al., 2003), human

1456 Neuron 110, May 4, 2022

 ll
Review

neuroimaging work has shown that specific patterns of activa- DYNAMIC CHANGES OF MEMORY UNDER STRESS
tion within the amygdala persist during ‘‘offline’’ periods shortly
following learning (Hermans et al., 2017). Furthermore, phenom- For long, stress research focused almost exclusively on stress-
ena of sequential reactivation (‘‘awake replay’’) of hippocampal induced changes in hippocampal spatial or declarative memory
neurons are potentiated following salient learning experiences formation or retrieval, and it was assumed that non-hippocam-
such as novel or rewarding events (Singer and Frank, 2009), pal memory would not be influenced by stress (Lupien et al.,
but also fear learning (Wu et al., 2017). 1997; Newcomer et al., 1994). This assumption, however, has
Phenomena of preferential re-instatement of learning-related been challenged by findings showing that stress and stress
activation patterns is not limited to single regions. For instance, hormones can affect memories that are independent of the hip-
increased synchronized theta-band oscillations were observed pocampus such as dorsal-striatum-based stimulus-response
between lateral amygdala and CA1 hippocampal region after memories. Systemic stress hormone administration or cortico-
fear learning (Seidenbecher et al., 2003), and coupling between steroid injection directly into the dorsal striatum affects these
lateral amygdala and CA1 increased following immobilization non-hippocampal memories in a similar manner as hippocam-
stress (Ghosh et al., 2013). In humans, functional connectivity be- pal memory, again enhancing the consolidation and impairing
tween amygdala and hippocampus measured using BOLD-fMRI the retrieval of these memories (Guenzel et al., 2013; Medina
was increased following fear learning, and this increase was et al., 2007).
associated with stronger fear memories (Hermans et al., 2017). Even more importantly, research over the past decade
Task-independent intrinsic functional connectivity between demonstrated that stress does not only result in quantitative
amygdala and hippocampus, measured at baseline, was further- changes in the performance of a single hippocampal or non-hip-
more predictive of later stress effects on declarative memory (de pocampal memory system but also in the balance between
Voogd et al., 2016), and categorical fear learning was shown to anatomically and functionally distinct memory systems (Packard
result in preferential reinstatement of neocortical representations and Goodman, 2012; Vogel et al., 2016). Often, multiple memory
of fear-associated semantic categories (de Voogd, 2016). It has systems are active at the same time that differ in the information
been proposed that the amygdala ‘‘gates’’ hippocampal-neocor- processed and may support different behavioral responses
tical communication by controlling the entorhinal-perirhinal (McDonald and White, 1993). Although highly relevant in stress-
pathway (Bauer et al., 2007), suggesting that the amygdala and ful situations, the differential contributions of these different
specifically stress-related noradrenergic activation plays a crit- memory systems could hardly be separated in tasks that were
ical role in permitting selective reactivations of memory represen- commonly used in previous research on stress and memory.
tations and hippocampal-neocortical crosstalk. This changed only when more complex learning tasks were em-
These noradrenergic effects may be complemented by func- ployed. Closely related to the stress-induced reconfiguration of
tionally synergistic corticosteroid actions. For instance, cortico- large-scale neural networks described above, accumulating ev-
steroids have been shown to gradually shift dominance of idence now suggests that stress determines which of these mul-
functional connectivity of the amygdala away from the SN and tiple memory systems governs behavior. More specifically, it has
toward regions involved in the DMN (Henckens et al., 2012), been demonstrated across tasks and species that stress or
which may prevent an overactivity that could be damaging if corticosteroid administration before learning induces a rapid
not controlled for. Corticosteroids are further critically implicated shift from reflective ‘‘cognitive’’ memory systems, such as the
in the upregulation of another large-scale network, the executive hippocampus or PFC, to more reflexive ‘‘habit’’ systems, such
control network (ECN), in the late phase of the stress response, as the amygdala or dorsolateral striatum (Kim et al., 2001;
i.e., >1 h after stress onset. This network, which supports higher- Schwabe et al., 2007; Siller-Pérez et al., 2017; Simon-Kutscher
order cognitive functions such as working memory, involves et al., 2019; Vogel et al., 2017; Wirz et al., 2018).
more dorsal prefrontal areas (dorsolateral PFC [dlPFC], precen- Converging lines of evidence from pharmacological and
tral/superior frontal sulci, and dorsomedial PFC) as well as pos- behavioral genetics studies suggest that this initial shift toward
terior parietal areas (Hermans et al., 2014a; Vincent et al., 2008). ‘‘habit’’ memory under stress is operated by non-genomic corti-
In line with rodent work showing that acute stress induces after costeroid action via the MR (Schwabe et al., 2010, 2013; Wirz
>1 h a long-lasting GR-dependent potentiation of excitatory et al., 2017a), presumably in close interaction with noradrenergic
neurotransmission in PFC (Yuen et al., 2009), human research activity (Packard and Goodman, 2012; Wirz et al., 2017b), while
has shown that administration of hydrocortisone exerts delayed the consequent consolidation of striatal memory depends on the
positive effects on PFC function (Henckens et al., 2011). This GR (Siller-Pérez et al., 2017). Notably, this shift from ‘‘cognitive’’
time delay of several hours is consistent with the temporal win- toward ‘‘habit’’ memory is not only observed during initial mem-
dow of potential genomic effects of corticosteroids. These ex- ory formation but also at retrieval (Elliott and Packard, 2008;
amples of a gradual ‘‘counterregulation’’ by corticosteroids in Zerbes et al., 2020; Zerbes and Schwabe, 2021). Thus, if multiple
DMN and ECN in the aftermath of a stressful event represents ‘‘cognitive’’ and more ‘‘habitual’’ memory traces exist in parallel,
a clear example of the complementary effects of quick noradren- acute stress leads to the predominance of habitual memory
ergic versus delayed corticosteroid activity (Hermans et al., retrieval, allowing well-established routines to guide behavior
2014a). It further suggests an active role for slow effects of under stress. We assume that the stress-induced bias from
corticosteroids in promoting consolidation and integration of in- ‘‘cognitive’’ to ‘‘habit’’ memory is a direct consequence of the
formation encoded during the acute phase of the stress neural network shift toward the SN, which includes, among other
response to promote behavioral adaptation. regions, the amygdala and dorsal striatum.

Neuron 110, May 4, 2022 1457

 ll
Review

Building directly on the stress-induced shift toward ‘‘habit’’ ination task affects the specificity of memory tested 28 days later
memory, recent research asked whether stress may impact— (Atucha et al., 2017). Results indicated that, compared with sa-
beyond the known effects on consolidation and retrieval—the line-treated rats that showed the expected transformation to
flexibility of memory. A key feature of adaptive memory is its ca- gist-like memory, memory remained detailed and specific at
pacity to flexibly guide future retrieval and hence behavior (Shoh- the 28-day retention test in NA-treated rats. Strikingly, this main-
amy and Adcock, 2010). Recent findings suggest that stress tenance of memory specificity after NA treatment was not only
hampers this mnemonic flexibility. For instance, stressed partic- associated with a maintenance of hippocampal dependency
ipants who were trained in a virtual navigation task showed an over time but even with increased hippocampal dependency,
increased reliance on familiar paths and reduced traversal of accompanied by changed patterns of DNA methylation and
shortcuts when these became available. Neuroimaging data re- mRNA expression of memory-related genes in the hippocampus
vealed that this deficit in flexible retrieval enabling efficient nav- and neocortex after 28 days, suggesting that NA may not only
igation was linked to reduced neural replay of memory for future slow down but even reverse systems consolidation. This pattern
locations and reduced activity relevant for mental simulation dur- of results was replicated in a very recent neuroimaging study in
ing probe trials (Brown et al., 2020). These findings dovetail with humans (Krenz et al., 2021). Here, increased noradrenergic ac-
recent evidence suggesting that stress may interfere with the ca- tivity shortly after encoding of pictures reduced the time-depen-
pacity to flexibly and intentionally control memory retrieval pro- dent decline in memory at a 28-day delayed test (relative to a 1-
cesses (Quaedflieg et al., 2020). Likewise, stress shortly before day delayed test) compared with placebo. In line with the rodent
initial learning or pharmacological elevations of noradrenergic data, fMRI findings showed that the noradrenergic stimulation
activity have been shown to impair participants’ ability to gener- led even to a time-dependent increase in hippocampal activity
alize across past experiences when required to flexibly transfer and episodic reinstatement during retention testing, accompa-
memories to novel situations (Dandolo and Schwabe, 2016; nied by a time-dependent decrease in neocortical activity.
Kluen et al., 2017a). This stress-related impairment in memory While these findings point to a critical impact of NA on the dy-
flexibility appears to extend to the ability to link existing mem- namics of memory formation and retrieval over time, another
ories with new information. Specifically, stress, NA, or cortico- question relates to the potential role of corticosteroids in the
steroids (administered shortly before training) impaired the long-lived changes in memory quality. Interestingly, there is
efficient use of existing knowledge to support new learning of initial evidence to suggest that NA and corticosteroids might
related material (Kluen et al., 2017b; Vogel et al., 2018) as well play complementary, or even opposite, roles in the dynamics
as the flexible updating of established memories in light of new of memory, underlining the idea that different stress mediators
information (Nitschke et al., 2019; Raio et al., 2017). These im- may have distinct roles in memory formation (see above):
pairments in mnemonic flexibility regarding incorporation of whereas post-encoding noradrenergic stimulation enhanced
new information may be closely linked to the reduced recruit- both memory strength and memory accuracy in the long run,
ment of the DMN and ECN under stress, presumably driven by most likely by increasing long-term hippocampal involvement
NA and rapid, non-genomic corticosteroid action via the MR. in memory, corticosterone led to strong but more generalized
Initial evidence suggests that these impairments in mnemonic memories, presumably by enhancing neocortical storage (Roo-
flexibility may be primarily owing to impaired flexibility of memory zendaal and Mirone, 2020). The finding that NA reinforces
retrieval (Quaedflieg et al., 2020; Vogel et al., 2018; Zerbes et al., episodic-like accuracy is consistent with other findings indi-
2020). However, this conclusion might be premature because cating that posttraining NA administration into the BLA enhances
participants were exposed to stress shortly before training in the accuracy of the association of an object with its specific
many studies which complicates a distinction between effects training context in an object-in-context recognition task (Bar-
on initial acquisition and subsequent retrieval processes. segyan et al., 2014) and maintains long-term accuracy of the
Furthermore, probes of memory flexibility typically involve the shock-context association on the inhibitory avoidance discrimi-
processing of new information against the background of prior nation task (Atucha et al., 2017). The finding that corticosterone
knowledge, i.e., a close interplay between acquisition and induces a generalized strengthening of memory is in agreement
retrieval processes. with previous evidence indicating that posttraining corticoste-
Another line of recent research asked how stress hormones rone administration also induced a generalization of fear memory
affect the long-term fate of memories. During systems consoli- and increased the freezing response to an innocuous auditory
dation, initially hippocampus-dependent memories are thought stimulus (Kaouane et al., 2012). Moreover, Dos Santos Corrêa
to become increasingly reliant on neocortical areas (Squire and et al. (2019) recently showed that a higher shock intensity during
Alvarez, 1995). This time-dependent reorganization is assumed contextual fear conditioning was associated with an enhanced
to be accompanied by a transformation from a detailed, episodic freezing response to a novel context, and that this generalization
memory trace to a more gist-like memory representation (Dan- effect positively correlated with corticosterone levels during the
dolo and Schwabe, 2018; Moscovitch and Gilboa, 2021). post-learning consolidation period.
Although this transformation may be generally adaptive to build In sum, research over the past two decades showed that acute
up abstract semantic knowledge structures, maintaining specific stress does not result in a global memory impairment—a view
and vivid memories over time may be particularly relevant for held by some decades ago—but rather that stress impairs
emotionally arousing or stressful events (Bahtiyar et al., 2020). some memory processes while enhancing others, critically
A recent study in rats tested whether NA administration into dependent on the exact timing of learning and retention testing
the BLA shortly after training on an inhibitory avoidance discrim- relative to the temporal profile of action of major stress mediators

1458 Neuron 110, May 4, 2022

 ll
Review

Figure 4. Memory changes under stress

Acute stress induces time-dependent changes in memory, enhancing some processes (green) while impairing others (red). These time-dependent changes in
memory are thought to be directly linked to the temporal profiles of action of major stress mediators (see Figure 1). Memory for key features of the stressful event
itself is typically enhanced. Further, stress may facilitate habitual forms of learning and memory. At the same time, stress can impair the formation and retrieval of
stressor-unrelated information as well as memory flexibility, as reflected in reduced goal-directed learning, impaired memory updating and hampered transfer of
memories to new situations. Both, the enhancing and impairing effects of stress are driven by rapidly acting catecholamines and corticosteroids, presumably in
interaction with other mediators such as the endocannabinoid system (but also peptides and other monoamines). Delayed, genomic corticosteroid actions,
however, may increase the threshold for encoding new information. This transient impairment in new memory formation might shield the consolidation of the
stressful event from interference and thus contribute to the recently shown long-term specificity of memories for arousing events.

and the experimental paradigm (Figure 4). Beyond the consoli- TOWARD AN INTEGRATIVE FRAMEWORK OF MEMORY
dation and retrieval of hippocampal memory, stress has been UNDER STRESS
shown to modulate non-hippocampal forms of memory as well
as the balancing of multiple, functionally distinct memory sys- The recent progress in our understanding of the stress-memory
tems. Most recent research further revealed that stress media- link at the cellular, neural network, and cognitive levels that we
tors may impair the flexible use and modification of memories have discussed in the preceding sections allows us to propose
but enhance the long-term specificity of memory (unless the sit- an integrative framework of how stress shapes memory.
uation is extremely stressful; Dos Santos Corrêa et al., 2019), This framework assumes that specific, time-dependent neural
with different stress hormones playing different roles in the latter. network shifts during and after stressful events represent an
We assume that these various effects of stress on memory pro- interface linking the orchestrated activity of multiple stress me-
cesses represent different shades of a common mechanism diators at the molecular and cellular level within areas, with
characterized by the time-dependent interplay of multiple stress distinct—but presumably interdependent—stress effects on
mediators and associated shifts in neural network balance. The the flexibility and long-term dynamics of memories at the
exact nature of the stress-induced changes in memory may cognitive level (Figure 5). Closely related to these time-depen-
further depend on the specific hormones that are released, influ- dent effects of stressors are potentially distinct roles of cate-
enced in part by the specific learning task but also individual cholamines and corticosteroids both in the shift between (areas
characteristics. belonging to) neural networks (Hermans et al., 2011; Van

Neuron 110, May 4, 2022 1459

 ll
Review

Figure 5. Integrative framework of how acute stress alters memory processes

Top: at the cellular level, rapid, non-genomic actions via MR in hippocampus (HC) and amygdala (AM) cells promote, directly or indirectly (e.g., through disin-
hibition), excitatory transmission. This is subsequently complemented by delayed, gene-mediated, and GR-dependent changes in cellular function that increase
signal-to-noise ratio in higher brain areas, such as hippocampus and prefrontal cortex (PFC).
Middle: these time-dependent cellular changes, linked to the specific temporal profiles of action of major stress mediators (see Figure 1), trigger time-dependent
reconfigurations of large-scale networks. During and shortly after a stressful event, when catecholaminergic and rapid corticosteroid actions prevail, there is a
shift toward a salience network (SN), at the expense of the executive control network (ECN) and default-mode network (DMN). At later stages, when cate-
cholamine effects have vanished and slow, genomic corticosteroid actions have developed, the network reconfiguration reverses.
Bottom: these network shifts translate directly into time-dependent changes in memory processes. The predominance of the salience network aids memory
formation for the stressful event, at the expense of the flexibility of memory and other processes, such as working memory or memory retrieval. The delayed shift
toward the ECN and DMN, transiently reduces memory formation but enhances mnemonic flexibility which might help to rationalize the stressful encounter. Both
the initial enhancement of memory formation and the delayed impairment of memory formation for new information, which may shield the memory formation for
the stressor itself, are assumed to contribute to the long-term specificity of memories for stressful or arousing events.

Stegeren et al., 2010) and in memory processes (Roozendaal creases neuronal activity in the amygdala (Atsak et al., 2015;
and Mirone, 2020). Campolongo et al., 2009). It is well established that the
During a stressful event, the rapid release of catecholamines amygdala can modulate memory processes in other brain
from brainstem nuclei is thought to have a circuit-breaking func- areas such as the hippocampus or dorsal striatum (McGaugh,
tion (Corbetta et al., 2008). It induces an increase in neural excit- 2015). Beyond these modulatory influences on single brain
ability of regions constituting the SN, which sets the stage for a areas, the amygdala has been shown to be critically implicated
network reconfiguration (Sara and Bouret, 2012). Key to this in orchestrating the shift from ‘‘cognitive’’ to ‘‘habit’’ memory
neural reconfiguration under stress is the amygdala, which inte- systems under stress (Kim et al., 2001; Schwabe et al., 2013;
grates the action of multiple stress mediators. Within the amyg- Vogel et al., 2017), presumably governed by non-genomic
dala, non-genomic corticosteroid actions via near-membrane corticosteroid actions via the MR and closely related to
receptors amplify NA effects on neuronal activity (Karst and the large-scale neural network reconfiguration directly after
Joe€ls, 2016; McGaugh, 2015; Roozendaal et al., 2004, 2006). stress from the ECN and DMN to the SN, specialized in pro-
In addition, corticosteroids trigger the release of ECBs, which cessing emotionally arousing events (Hermans et al., 2011,
then bind to CB1 receptors on GABAergic interneurons to inhibit 2014b). These small- and large-scale neural network changes,
GABA release (Di et al., 2016). This, in turn, may disinhibit the mainly driven by the rapid effects of multiple stress mediators
release of NA from presynaptic sites and hence further in- on both excitatory and inhibitory neurons, form the basis

1460 Neuron 110, May 4, 2022

 ll
Review

for the manifold changes in mnemonic processes that are in the late phase of the stress response, now favoring the DMN
observed during and shortly after stress. Specifically, the and ECN over the SN (Hermans et al., 2014b). Indeed, PFC func-
amygdala-induced modulation of hippocampal activity and tions appear to be enhanced when genomic corticosteroid ac-
plasticity—presumably paralleled by facilitating effects of rapid tions are active (Henckens et al., 2011; Yuen et al., 2009),
corticosteroid actions directly in the hippocampus (Wiegert whereas emotional reactivity closely related to the SN is reduced
et al., 2006)—results in the enhancement of memory formation (Putman et al., 2007). This delayed network reversal may thus not
for the stressful event itself (Kalbe et al., 2020; Sandi et al., only help the organism to restore homeostasis but also to ratio-
1997; Vogel and Schwabe, 2016). Notably, this memory nalize, contextualize, and store the stressful experience into
enhancement is only observed for information directly relevant long-term memory.
to the ongoing stressor, whereas the encoding of information Importantly, the accumulating evidence suggesting that
that is present during the stressful episode but not directly different stress mediators, in particular catecholamines and cor-
relevant to stressor is even reduced (Kalbe et al., 2020; ticosteroids, may play distinct roles in neural network changes
Schwabe and Wolf, 2010). The boost in memory formation for and long-term memory specificity (Hermans et al., 2014b; Roo-
the stressful event itself may be driven by the shift toward the zendaal and Mirone, 2020; Van Stegeren et al., 2010) indicates
SN, which involves also sensory representation areas, known that the specific effects of a stressful experience on memory
to interact with the hippocampus in forming long-lasting mem- depend on the specific endocrine stress response pattern, which
ories of stressful events (de Voogd, 2016). The large-scale shift may differ across individuals and types of stressors.
toward the SN further promotes the predominance of habits and
routines that are frequently observed under stress and rely on CLINICAL IMPLICATIONS
striatal areas belonging to the SN (Vogel et al., 2016; Wirz
et al., 2018). Given that several mental disorders are characterized by altered
This bias toward the SN, however, may come at the cost of the stress response patterns and that stress-induced changes in
ECN and DMN, including medial and lateral PFC. These prefron- memory are thought to be a driving force in stress-related mental
tal areas are crucial for flexible, goal-directed behavior (Balleine disorders, research on the impact of stress on memory comes
and O’Doherty, 2010). Moreover, memory retrieval processes, with the hope that it will promote our understanding of these
the transfer and generalization of memories, memory updating disorders and might ultimately lead to novel treatment ap-
as well as learning against the background of existing knowl- proaches. Indeed, several interventions that build directly on
edge, all of these processes are heavily dependent on the PFC basic research on the stress-memory link have been suggested.
and its crosstalk with MTL areas (Preston and Eichenbaum, For instance, pharmacological treatments targeting corticoste-
2013; Shin et al., 2019). Thus, both the impairment of memory roid or NA signaling were used to either facilitate the consolida-
retrieval and the reduced mnemonic flexibility under stress tion of therapeutic interventions or interfere with the retrieval of
may be due to the large-scale neural reconfiguration at the dysfunctional memories in phobia or PTSD (for a review see de
expense of the ECN and DMN. At the same time, the downregu- Quervain et al., 2017), and more lately in addiction (Soravia
lation of prefrontal and parietal storage sites may lay the ground et al., 2021). Beyond pharmacological treatments, cognitive in-
for an altered systems consolidation process. Combined with terventions were recently introduced that aim at either contextu-
specific synaptic changes within the hippocampus after alizing or intentionally controlling overly strong memories for
increased noradrenergic stimulation, altered communication be- highly stressful, traumatic events (Abed et al., 2020; Mary
tween the hippocampus and neocortical storage sites is thought et al., 2020).
to contribute to the subsequent increase in long-term specificity The recently discovered changes in memory under stress that
of and increased hippocampal involvement in remote memories we have discussed here may enhance our understanding of
(Atucha et al., 2017; Krenz et al., 2021). Thus, we assume that the mental disorders such as PTSD. For instance, the long-term
transiently impaired flexibility of memory and its long-term spec- specificity of memory due to increased noradrenergic activity
ificity reflect two sides of the same coin, both being due the shortly after encoding (Atucha et al., 2017; Krenz et al., 2021)
downregulation of the ECN and DMN during and shortly after a may contribute to the vividness and longevity of trauma memory.
stressful experience. These network changes, in turn, are Furthermore, the transient decrease in memory flexibility may
thought to be driven by rapid catecholamine and non-genomic result in rather rigid memories (Wirz et al., 2018) that lack contex-
corticosteroid actions. tual details (Simon-Kutscher et al., 2019; van Ast et al., 2013).
Once the stressful event is over, catecholamine effects vanish Such rigid memories could explain the overly strong emotional
rapidly and genomic corticosteroid actions develop within 1–2 h responding to single trauma-related cues (e.g., odors and tones)
after stressor onset (Joe €ls and Baram, 2009). These delayed in PTSD patients and may complicate therapeutic interventions.
stress effects are assumed to reduce hippocampal and amygda- There is further recent evidence that directly links an impairment
lar neuroplasticity related to the encoding of new information of the flexible control of memory retrieval, as observed under
(Diamond et al., 2007; Joe €ls et al., 2006), shifting the organism acute stress (Quaedflieg et al., 2020), to PTSD symptoms (Cata-
to a ‘‘memory storage mode’’ (Roozendaal, 2002; Schwabe rino et al., 2015; Mary et al., 2020).
et al., 2012) that further protects the consolidation of the stressful Beyond the enhanced understanding of the potential mecha-
event from interference and allows the synaptic reorganization nisms contributing to stress-related mental disorders, several
required for the long-term specificity of memory. Moreover, the specific routes for intervention can be directly derived from the
large-scale network reconfiguration is assumed to be reversed mechanistic framework that we propose here. First, in light of

Neuron 110, May 4, 2022 1461

 ll
Review

the recently identified role of ECBs in stress effects on memory, level with time-dependent large-scale network shifts at the sys-
potential pharmacological interventions targeting both cortico- tems level and specific changes at the cognitive level, which
steroid and ECB signaling might be particularly promising (Neu- restrict the possibilities to flexibly update and readjust memories
meister et al., 2013), presumably in combination with exposure but may pave the way for long-lasting enhancements of memory
therapy. Second, as we have argued above, large-scale neural specificity.
network changes may be the driving force in stress-induced The progress in research on stress and memory that we have
changes in memory, and hence, we assume that stress-related seen over the past decades was at least partly linked to
mental disorders originate from changes at the network level. emerging technical advances. For instance, optogenetics or ge-
Indeed, there is accumulating evidence suggesting such netic modifications in rodents are now providing tools to target
network changes in disorders such as PTSD (Fonzo et al., the mechanisms underlying stress-induced changes in memory
2021), which might derive from a disbalance in stress response with unprecedented precision. In humans, the investigation of
patterns. Here, recent evidence is of interest showing that trans- large-scale neural networks became only possible with the
cranial magnetic stimulation (TMS) over defined cortical sites advent of whole-brain fMRI. However, research on stress and
can be used to modify entire networks of interconnected brain memory has recently only begun to leverage the potential asso-
areas (Philip et al., 2018). Given the assumed importance of ciated with multivariate analyses of neuroimaging data, machine
the quick shift from the ECN and DMN to the SN early on in learning, or cognitive modeling to elucidate the neural and cogni-
stress-induced changes of memory, (repeated) TMS over tive mechanisms involved in memory under stress (Gagnon et al.,
cortical nodes of the ECN or DMN may be employed to prevent 2019; Lenow et al., 2017; Meier et al., 2021). These and the tech-
a prolonged or dysfunctional network reconfiguration under nical or methodological advances to come may allow us tackling
stress. Moreover, deficits in memory retrieval or flexibility under the fundamental questions related to the stress-memory link that
stress might be attenuated by rebalancing cortical excitation and remained unanswered so far.
inhibition, recently shown to be altered under acute stress (Han One key question for future research is how the different stress
et al., 2020), via transcranial direct current stimulation over response waves, such as rapid, non-genomic, and slow
cortical sites of the ECN or DMN (Barron et al., 2016; Koolschijn genomic corticosteroid actions, relate to one another, not only
et al., 2019). Finally, the identified mechanisms could help to at the cellular level but also at the integrated network level;
identify relevant (epi)genetic (Vukojevic et al., 2020) or neural how the waves interact and which prevails if they overlap.
(van Leeuwen et al., 2019) risk markers, e.g., related to ECB Closely related, it is not well understood exactly how the interac-
signaling, expression of stress hormone receptors or large-scale tion of multiple stress mediators results in a temporally dynamic
network balance, for individuals who are particularly vulnerable reconfiguration of large-scale networks. How can ECBs regulate
to maladaptive stress effects on memory. The individual vulner- stress-induced network shifts? Further, if the amygdala, as pro-
ability to stress-related disorders may be linked to different posed here, plays a critical role in this configuration, how does
stages of memory (formation versus retrieval) and to specific this process work? And how can the amygdala operate the
phases of the stress response (rapid versus delayed). Balancing switch between networks depending on the prevailing stress
the different stress response phases to avoid either overshooting response mode? In general, while many studies focused on
or failing memory formation for the stressful event appears to be the effects of stress around encoding or retrieval on subsequent
crucial to prevent aberrant stress effects on memory and ulti- memory, less is known about the delayed effects of stress on
mately protect mental health. memory formation or retrieval and its neural basis, in particular
in humans (but see Henckens et al., 2010, 2011). Moreover,
CONCLUSIONS AND FUTURE DIRECTIONS although recent findings show that stress hormones can bidirec-
tionally modulate systems consolidation processes and influ-
We have provided an inevitably selective review of recent ad- ence memory specificity weeks to months later (Atucha et al.,
vances in our understanding of the mechanisms through which 2017; Krenz et al., 2021; Roozendaal and Mirone, 2020), which
stressful events shape memory. Recent evidence indicates an cellular and network mechanisms enable such long-lasting ef-
interaction between multiple stress mediators, in which each of fects remains completely unknown. Finally, a major challenge
these mediators appears to play a distinctive role, with comple- for future research relates to how existing ideas about the role
mentary or sometimes even opposite effects of major stress me- of stress and memory in the development of mental disorders
diators, such as NA and corticosteroids, depending on the time now based largely on findings in healthy humans can be put to
after stress and the receptors involved. It is further becoming test in specific patient populations and be ultimately translated
increasingly clear that stress effects on memory cannot be un- into (personalized) interventions for disorders characterized by
derstood at the level of isolated cells or even brain areas, such maladaptive changes in memory under stress.
as the hippocampus or the amygdala, despite the fact that these
have been useful for a guiding theoretical framework, but that ACKNOWLEDGMENTS
these effects rely on complete microcircuits and shifts in large-
scale neural networks. These network shifts translate into behav- L.S. received funding by the Deutsche Forschungsgemeinschaft (DFG,
ioral and cognitive changes that are much more complex and German Research Foundation)—178316478 – B10. E.J.H. is supported by
longer-lasting than previously thought. Based on these ad- the European Research Council (ERC-2015-CoG 682591). B.R. is supported
by the Netherlands Organization for Scientific Research (NWO-ORA
vances, we propose an integrative framework that links the 464.18.1100). We gratefully acknowledge the helpful comments of Ron de
orchestrated action of multiple stress mediators at the cellular Kloet on an earlier version of this manuscript.

1462 Neuron 110, May 4, 2022

 ll
Review
AUTHOR CONTRIBUTIONS facilitation of prelimbic cortex interactions. Proc. Natl. Acad. Sci. USA 116,
7077–7082.
All authors contributed to the conceptualization, writing, and editing of the Bauer, E.P., Paz, R., and Paré, D. (2007). Gamma oscillations coordinate
manuscript. amygdalo-rhinal interactions during learning. J. Neurosci. 27, 9369–9379.

DECLARATION OF INTERESTS Bonapersona, V., Schuler, H., Damsteegt, R.D., Adolfs, Y., Pasterkamp, R.J.,
van den Heuvel, M.P., Joe €ls, M., and Sarabdjitsingh, R.A. (2022). The mouse
brain after foot-shock in 4D: temporal dynamics at a single-cell resolution.
The authors declare no competing interests. Proc. Natl. Acad. Sci. USA 119, e2114002119.

Bouarab, C., Roullot-Lacarrière, V., Vallée, M., Le Roux, A., Guette, C., Men-
REFERENCES
nesson, M., Marighetto, A., Desmedt, A., Piazza, P.V., and Revest, J.M.
(2021). PAI-1 protein is a key molecular effector in the transition from normal
Adamsky, A., Kol, A., Kreisel, T., Doron, A., Ozeri-Engelhard, N., Melcer, T., to PTSD-like fear memory. Mol. Psychiatry 26, 4968–4981. https://doi.org/
Refaeli, R., Horn, H., Regev, L., Groysman, M., et al. (2018). Astrocytic activa- 10.1038/s41380-021-01024-1.
tion generates de novo neuronal potentiation and memory enhancement. Cell
174, 59–71.e14. Brown, T.I., Gagnon, S.A., and Wagner, A.D. (2020). Stress disrupts human
hippocampal-prefrontal function during prospective spatial navigation and
Akirav, I., Kozenicky, M., Tal, D., Sandi, C., Venero, C., and Richter-Levin, G. hinders flexible behavior. Curr. Biol. 30, 1821–1833.e8.
(2004). A Facilitative Role for Corticosterone in the Acquisition of a Spatial
Task Under Moderate Stress. Learn Mem 11, 188–195. Buchanan, T.W., and Lovallo, W.R. (2001). Enhanced memory for emotional
material following stress-level cortisol treatment in humans. Psychoneuroen-
Al Abed, A.S., Ducourneau, E.G., Bouarab, C., Sellami, A., Marighetto, A., and docrinology 26, 307–317.
Desmedt, A. (2020). Preventing and treating PTSD-like memory by trauma
contextualization. Nat. Commun. 11, 4220. Buchanan, T.W., Tranel, D., and Adolphs, R. (2006). Impaired memory retrieval
correlates with individual differences in cortisol response but not autonomic
Andreano, J.M., and Cahill, L. (2009). Sex influences on the neurobiology of response. Learn Mem 13, 382–387.
learning and memory. Learn. Mem. 16, 248–266.
Buckner, R.L., Andrews-Hanna, J.R., and Schacter, D.L. (2008). The brain’s
Arnsten, A.F.T. (2009). Stress signalling pathways that impair prefrontal cortex default network: anatomy, function, and relevance to disease. Ann. N. Y.
structure and function. Nat. Rev. Neurosci. 10, 410–422. Acad. Sci. 1124, 1–38.

Atsak, P., Hauer, D., Campolongo, P., Schelling, G., Fornari, R.V., and Roozen- Buurstede, J.C., van Weert, L.T.C.M., Colucci, P., Gentenaar, M., Viho,
daal, B. (2015). Endocannabinoid signaling within the basolateral amygdala E.M.G., Koorneef, L.L., Schoonderwoerd, R.A., Lanooij, S.D., Moustakas, I.,
integrates multiple stress hormone effects on memory consolidation. Neuro- Balog, J., et al. (2021). Hippocampal glucocorticoid target genes associated
psychopharmacology 40, 1485–1494. with enhancement of memory consolidation. Eur. J. Neurosci. https://doi.
org/10.1111/ejn.15226.
Atsak, P., Hauer, D., Campolongo, P., Schelling, G., McGaugh, J.L., and Roo-
zendaal, B. (2012). Glucocorticoids interact with the hippocampal endocanna- Cahill, L., Gorski, L., and Le, K. (2003). Enhanced human memory consolida-
binoid system in impairing retrieval of contextual fear memory. Proc. Natl. tion with post-learning stress: interaction with the degree of arousal at encod-
Acad. Sci. USA 109, 3504–3509. ing. Learn. Mem. 10, 270–274.

Atucha, E., Vukojevic, V., Fornari, R.V., Ronzoni, G., Demougin, P., Peter, F., Campolongo, P., Roozendaal, B., Trezza, V., Hauer, D., Schelling, G.,
Atsak, P., Coolen, M.W., Papassotiropoulos, A., McGaugh, J.L., et al. McGaugh, J.L., and Cuomo, V. (2009). Endocannabinoids in the rat basolateral
(2017). Noradrenergic activation of the basolateral amygdala maintains hippo- amygdala enhance memory consolidation and enable glucocorticoid modula-
campus-dependent accuracy of remote memory. Proc. Natl. Acad. Sci. USA tion of memory. Proc. Natl. Acad. Sci. USA 106, 4888–4893.
114, 9176–9181.
€pper, C.S., Werner-Seidler, A., Dalgleish, T., and Anderson,
Catarino, A., Ku
Bahtiyar, S., Gulmez Karaca, K.G., Henckens, M.J.A.G., and Roozendaal, B. M.C. (2015). Failing to forget: inhibitory-control deficits compromise memory
(2020). Norepinephrine and glucocorticoid effects on the brain mechanisms suppression in posttraumatic stress disorder. Psychol. Sci. 26, 604–616.
underlying memory accuracy and generalization. Mol. Cell. Neurosci. 108,
103537. Chen, Y., Barsegyan, A., Nadif Kasri, N., and Roozendaal, B. (2018). Basolat-
eral amygdala noradrenergic activity is required for enhancement of object
Bains, J.S., Wamsteeker Cusulin, J.I., and Inoue, W. (2015). Stress-related recognition memory by histone deacetylase inhibition in the anterior insular
synaptic plasticity in the hypothalamus. Nat. Rev. Neurosci. 16, 377–388. cortex. Neuropharmacology 141, 32–41.

Balleine, B.W., and O’Doherty, J.P. (2010). Human and rodent homologies in Clayton, D.C., Anreiter, I., Aristizabal, M., Frankland, P.W., Binder, E.B., and
action control: corticostriatal determinants of goal-directed and habitual ac- Citri, A. (2020). The role of the genome in experience-dependent plasticity: Ex-
tion. Neuropsychopharmacology 35, 48–69. tending the analogy of the genomic action potential. Proc Natl Acad Sci U S A
117, 12252–23260.
Bar, M. (2021). From objects to unified minds. Curr. Dir. Psychol. Sci. 30,
129–137. Conrad, C.D. (2010). A critical review of chronic stress effects on spatial
learning and memory. Prog. Neuropsychopharmacol. Biol. Psychiatry 34,
Barron, H.C., Vogels, T.P., Emir, U.E., Makin, T.R., O’Shea, J., Clare, S., 742–755.
Jbabdi, S., Dolan, R.J., and Behrens, T.E.J. (2016). Unmasking latent inhibitory
connections in human cortex to reveal dormant cortical memories. Neuron 90, Corbetta, M., Patel, G., and Shulman, G.L. (2008). The reorienting system of
191–203. the human brain: from environment to theory of mind. Neuron 58, 306–324.

Barsegyan, A., Mackenzie, S.M., Kurose, B.D., McGaugh, J.L., and Roozen- Dallman, M.F. (2005). Fast glucocorticoid actions on brain: back to the future.
daal, B. (2010). Glucocorticoids in the prefrontal cortex enhance memory Front. Neuroendocrinol. 26, 103–108.
consolidation and impair working memory by a common neural mechanism.
Proc. Natl. Acad. Sci. USA 107, 16655–16660. Dandolo, L.C., and Schwabe, L. (2016). Stress-induced cortisol hampers
memory generalization. Learn. Mem. 23, 679–683.
Barsegyan, A., McGaugh, J.L., and Roozendaal, B. (2014). Noradrenergic acti-
vation of the basolateral amygdala modulates the consolidation of object-in- Dandolo, L.C., and Schwabe, L. (2018). Time-dependent memory transforma-
context recognition memory. Front. Behav. Neurosci. 8, 160. tion along the hippocampal anterior-posterior axis. Nature Communications
9, 1205.
Barsegyan, A., Mirone, G., Ronzoni, G., Guo, C., Song, Q., van Kuppeveld, D.,
Schut, E.H.S., Atsak, P., Teurlings, S., McGaugh, J.L., et al. (2019). Glucocor- €ls, M. (1999). Stress and cognition: are cor-
de Kloet, E.R., Oitzl, M.S., and Joe
ticoid enhancement of recognition memory via basolateral amygdala-driven ticosteroids good or bad guys? Trends Neurosci 22, 422–426.

Neuron 110, May 4, 2022 1463

 ll
Review
de Quervain, D.J., Roozendaal, B., and McGaugh, J.L. (1998). Stress and glu- Hamann, S.B., Ely, T.D., Grafton, S.T., and Kilts, C.D. (1999). Amygdala activity
cocorticoids impair retrieval of long-term spatial memory. Nature 394, related to enhanced memory for pleasant and aversive stimuli. Nat. Neurosci.
787–790. 2, 289–293.

de Quervain, D.J., Roozendaal, B., Nitsch, R.M., McGaugh, J.L., and Hock, C. Han, K., Lee, M., Lim, H.-K., Jang, M.W., Kwon, J., Lee, C.J., Kim, S.-G., and
(2000). Acute cortisone administration impairs retrieval of long-term declara- Suh, M. (2020). Excitation-inhibition imbalance leads to alteration of neuronal
tive memory in humans. Nat Neurosci 3, 313–314. coherence and neurovascular coupling under acute stress. J. Neurosci. 40,
9148–9162.
de Quervain, D., Schwabe, L., and Roozendaal, B. (2017). Stress, glucocorti-
coids and memory: implications for treating fear-related disorders. Nat. Rev. Hartmann, J., Dedic, N., Pöhlmann, M.L., Ha€usl, A., Karst, H., Engelhardt, C.,
Neurosci. 18, 7–19. Westerholz, S., Wagner, K.V., Labermaier, C., Hoeijmakers, L., et al. (2017).
Forebrain glutamatergic, but not GABAergic, neurons mediate anxiogenic ef-
de Voogd, L.D., Fernández, G., and Hermans, E.J. (2016). Awake reactivation fects of the glucocorticoid receptor. Mol. Psychiatry 22, 466–475.
of emotional memory traces through hippocampal-neocortical interactions.
Neuroimage 134, 563–572. Hatfield, T., Spanis, C., and McGaugh, J.L. (1999). Response of amygdalar
norepinephrine to footshock and GABAergic drugs using in vivo microdialysis
de Voogd, L.D., Klumpers, F., Fernández, G., and Hermans, E.J. (2017). and HPLC. Brain Res. 835, 340–345.
Intrinsic functional connectivity between amygdala and hippocampus during
rest predicts enhanced memory under stress. Psychoneuroendocrinology Hebert-Chatelain, E., Desprez, T., Serrat, R., Bellocchio, L., Soria-Gomez, E.,
75, 192–202. Busquets-Garcia, A., Pagano Zottola, A.C., Delamarre, A., Cannich, A., Vin-
cent, P., et al. (2016). A cannabinoid link between mitochondria and memory.
den Boon, F.S., de Vries, T., Baelde, M., Joe€ls, M., and Karst, H. (2019). Circa- Nature 539, 555–559.
dian and ultradian variations in corticosterone level influence functioning of the
male mouse basolateral amygdala. Endocrinology 160, 791–802. Henckens, M.J.A.G., Deussing, J.M., and Chen, A. (2016). Region-specific
roles of the corticotropin-releasing factor-urocortin system in stress. Nat.
Di, S., Itoga, C.A., Fisher, M.O., Solomonow, J., Roltsch, E.A., Gilpin, N.W., Rev. Neurosci. 17, 636–651.
and Tasker, J.G. (2016). Acute stress suppresses synaptic inhibition and in-
creases anxiety via endocannabinoid release in the basolateral amygdala. Henckens, M.J.A.G., van Wingen, G.A., Joe€ls, M., and Fernández, G. (2010).
J. Neurosci. 36, 8461–8470. Time-dependent effects of corticosteroids on human amygdala processing.
J. Neurosci. 30, 12725–12732.
Di, S., Malcher-Lopes, R., Halmos, K.C., and Tasker, J.G. (2003). Nongenomic
glucocorticoid inhibition via endocannabinoid release in the hypothalamus: a Henckens, M.J.A.G., Van Wingen, G.A., Joe €ls, M., and Fernández, G. (2011).
fast feedback mechanism. J. Neurosci. 23, 4850–4857. Time-dependent corticosteroid modulation of prefrontal working memory pro-
cessing. Proc. Natl. Acad. Sci. USA 108, 5801–5806.
Diamond, D.M., Campbell, A.M., Park, C.R., Halonen, J., and Zoladz, P.R.
(2007). The temporal dynamics model of emotional memory processing: a syn- Henckens, M.J.A.G., Van Wingen, G.A., Joe€ls, M., and Fernández, G. (2012).
thesis on the neurobiological basis of stress-induced amnesia, flashbulb and Corticosteroid induced decoupling of the amygdala in men. Cereb. Cortex
traumatic memories, and the Yerkes-Dodson law. Neural Plast. 2007, 60803. 22, 2336–2345.
Diamond, D.M., and Rose, G.M. (1994). Stress impairs LTP and hippocampal-
Hermans, E.J., Battaglia, F.P., Atsak, P., de Voogd, L.D., Fernández, G., and
dependent memory. Acad Sci 746, 411–414.
Roozendaal, B. (2014a). How the amygdala affects emotional memory by
altering brain network properties. Neurobiol. Learn. Mem. 112, 2–16.
Dos Santos Corrêa, M., dos Vaz, B.D.S., Grisanti, G.D.V., de Paiva, J.P.Q.,
Tiba, P.A., and Fornari, R.V. (2019). Relationship between footshock intensity, €ls, M., and Fernández, G. (2014b). Dy-
Hermans, E.J., Henckens, M.J.A.G., Joe
post-training corticosterone release and contextual fear memory specificity
namic adaptation of large-scale brain networks in response to acute stressors.
over time. Psychoneuroendocrinology 110, 104447.
Trends Neurosci. 37, 304–314.
Elliott, A.E., and Packard, M.G. (2008). Intra-amygdala anxiogenic drug infu-
sion prior to retrieval biases rats to the use of habit memory. Neurobiol. Learn. Hermans, E.J., Kanen, J.W., Tambini, A., Fernández, G., Davachi, L., and
Phelps, E.A. (2017). Persistence of amygdala-hippocampal connectivity and
Mem. 90, 616–623.
multi-voxel correlation structures during awake rest after fear learning predicts
Ferry, B., Roozendaal, B., and McGaugh, J.L. (1999). Basolateral amygdala long-term expression of fear. Cereb. Cortex 27, 3028–3041.
noradrenergic influences on memory storage are mediated by an interaction
between beta- and alpha1-adrenoceptors. J. Neurosci. 19, 5119–5123. Hermans, E.J., van Marle, H.J.F., Ossewaarde, L., Henckens, M.J.A.G., Qin,
S., van Kesteren, M.T.R., Schoots, V.C., Cousijn, H., Rijpkema, M., Oosten-
Filiou, M.D., and Sandi, C. (2019). Anxiety and brain mitochondria: a bidirec- veld, R., and Fernandez, G. (2011). Stress-related noradrenergic activity
tional crosstalk. Trends Neurosci. 42, 573–588. prompts large-scale neural network reconfiguration. Science 334, 1151–1153.

Fonzo, G.A., Goodkind, M.S., Oathes, D.J., Zaiko, Y.V., Harvey, M., Peng, K.K., Hill, M.N., Karatsoreos, I.N., Hillard, C.J., and McEwen, B.S. (2010). Rapid el-
Weiss, M.E., Thompson, A.L., Zack, S.E., Lindley, S.E., et al. (2021). Amygdala evations in limbic endocannabinoid content by glucocorticoid hormones
and insula connectivity changes following psychotherapy for posttraumatic in vivo. Psychoneuroendocrinology 35, 1333–1338.
stress disorder: a randomized clinical trial. Biol. Psychiatry 89, 857–867.
Hirschberg, S., Li, Y., Randall, A., Kremer, E.J., and Pickering, A.E. (2017).
Gagnon, S.A., Waskom, M.L., Brown, T.I., and Wagner, A.D. (2019). Stress im- Functional dichotomy in spinal- vs prefrontal-projecting locus coeruleus mod-
pairs memory retrieval by disrupting hippocampal and cortical mechanisms of ules splits descending noradrenergic analgesia from ascending aversion and
remembering. Cereb. Cortex 29, 2947–2964. anxiety in rats. Elife 6, e29808.

Ghosh, S., Laxmi, T.R., and Chattarji, S. (2013). Functional connectivity from Ikegaya, Y., Nakanishi, K., Saito, H., and Abe, K. (1997). Amygdala beta-
the amygdala to the hippocampus grows stronger after stress. J. Neurosci. noradrenergic influence on hippocampal long-term potentiation in vivo.
33, 7234–7244. NeuroReport 8, 3143–3146.

Gilboa, A., and Moscovitch, M. (2021). No consolidation without representa- Inoue, R., Abdou, K., Hayashi-Tanaka, A., Muramatsu, S.I., Mino, K., Inokuchi,
tion: correspondence between neural and psychological representations in K., and Mori, H. (2018). Glucocorticoid receptor-mediated amygdalar meta-
recent and remote memory. Neuron 109, 2239–2255. plasticity underlies adaptive modulation of fear memory by stress. Elife 7,
e34135.
Groc, L., Choquet, D., and Chaouloff, F. (2008). The stress hormone cortico-
sterone conditions AMPAR surface trafficking and synaptic potentiation. Ji, D., and Wilson, M.A. (2007). Coordinated memory replay in the visual cortex
Nat. Neurosci. 11, 868–870. and hippocampus during sleep. Nat. Neurosci. 10, 100–107.

Guenzel, F.M., Wolf, O.T., and Schwabe, L. (2013). Stress disrupts response €ls, M., and Baram, T.Z. (2009). The neuro-symphony of stress. Nat. Rev.
Joe
memory retrieval. Psychoneuroendocrinology 38, 1460–1465. Neurosci. 10, 459–466.

1464 Neuron 110, May 4, 2022

 ll
Review
€ls, M., Fernandez, G., and Roozendaal, B. (2011). Stress and emotional
Joe Lupien, S.J., Gaudreau, S., Tchiteya, B.M., Maheu, F., Sharma, S., Nair, N.P.,
memory: a matter of timing. Trends Cogn Sci 15, 280–286. Hauger, R.L., McEwen, B.S., and Meaney, M.J. (1997). Stress-induced declar-
ative memory impairment in healthy elderly subjects: relationship to cortisol
€ls, M., Karst, H., and Sarabdjitsingh, R.A. (2018). The stressed brain of hu-
Joe reactivity. J. Clin. Endocrinol. Metab. 82, 2070–2075.
mans and rodents. Acta Physiol. Oxf. 223, e13066.
Lupien, S.J., McEwen, B.S., Gunnar, M.R., and Heim, C. (2009). Effects of
€ls, M., Pu, Z., Wiegert, O., Oitzl, M.S., and Krugers, H.J. (2006). Learning
Joe stress throughout the lifespan on the brain, behavior and cognition. Nat.
under stress: how does it work? Trends Cogn. Sci. 10, 152–158. Rev. Neurosci. 10, 434–445.
€ls, M., Sarabdjitsingh, R.A., and Karst, H. (2012). Unraveling the time do-
Joe Mary, A., Dayan, J., Leone, G., Postel, C., Fraisse, F., Malle, C., Vallée, T.,
mains of corticosteroid hormone influences on brain activity: rapid, slow, Klein-Peschanski, C., Viader, F., de la Sayette, V., et al. (2020). Resilience after
and chronic modes. Pharmacol. Rev. 64, 901–938. trauma: the role of memory suppression. Science 367, 8477.
Kalbe, F., Bange, S., Lutz, A., and Schwabe, L. (2020). Expectancy violation McCall, J.G., Al-Hasani, R., Siuda, E.R., Hong, D.Y., Norris, A.J., Ford, C.P.,
drives memory boost for stressful events. Psychol. Sci. 31, 1409–1421. and Bruchas, M.R. (2015). CRH engagement of the locus coeruleus noradren-
ergic system mediates stress-induced anxiety. Neuron 87, 605–620.
Kaouane, N., Porte, Y., Vallée, M., Brayda-Bruno, L., Mons, N., Calandreau, L.,
Marighetto, A., Piazza, P.V., and Desmedt, A. (2012). Glucocorticoids can
McDonald, R.J., and White, N.M. (1993). A triple dissociation of memory sys-
induce PTSD-like memory impairments in mice. Science 335, 1510–1513.
tems: hippocampus, amygdala, and dorsal striatum. Behav. Neurosci.
€ls, M. (2010). Metaplas-
€tz, G., and Joe
Karst, H., Berger, S., Erdmann, G., Schu 107, 3–22.
ticity of amygdalar responses to the stress hormone corticosterone. Proc.
Natl. Acad. Sci. USA 107, 14449–14454. McEwen, B.S. (2002). The End of Stress as we Know it (Joseph Henry Press).

€ls, M. (2005).
€tz, G., and Joe
Karst, H., Berger, S., Turiault, M., Tronche, F., Schu McEwen, B.S., and Sapolsky, R.M. (1995). Stress and cognitive function. Curr.
Mineralocorticoid receptors are indispensable for nongenomic modulation of Opin. Neurobiol. 5, 205–216.
hippocampal glutamate transmission by corticosterone. Proc. Natl. Acad.
Sci. USA 102, 19204–19207. McEwen, B.S., Weiss, J.M., and Schwartz, L.S. (1968). Selective retention of
corticosterone by limbic structures in rat brain. Nature 220, 911–912.
Karst, H., and Joe€ls, M. (2005). Corticosterone slowly enhances miniature
excitatory postsynaptic current amplitude in mice CA1 hippocampal cells. McGaugh, J.L. (2000). Memory: a century of consolidation. Science 287,
J. Neurophysiol. 94, 3479–3486. 248–251.

€ls, M. (2016). Severe stress hormone conditions cause an

Karst, H., and Joe McGaugh, J.L. (2003). Memory and Emotion: the Making of Lasting Memories
extended window of excitability in the mouse basolateral amygdala. Neuro- (Columbia University Press).
pharmacology 110, 175–180.
McGaugh, J.L. (2015). Consolidating memories. Annu. Rev. Psychol. 66, 1–24.
Katona, I., Rancz, E.A., Acsady, L., Ledent, C., Mackie, K., Hajos, N., and
Freund, T.F. (2001). Distribution of CB1 cannabinoid receptors in the amygdala McIntyre, C.K., Miyashita, T., Setlow, B., Marjon, K.D., Steward, O., Guzowski,
and their role in the control of GABAergic transmission. J. Neurosci. 21, J.F., and McGaugh, J.L. (2005). Memory-influencing intra-basolateral amyg-
9506–9518. dala drug infusions modulate expression of Arc protein in the hippocampus.
Proc. Natl. Acad. Sci. USA 102, 10718–10723.
Kim, H. (2011). Neural activity that predicts subsequent memory and forget-
ting: a meta-analysis of 74 fMRI studies. Neuroimage 54, 2446–2461. McReynolds, J.R., Donowho, K., Abdi, A., McGaugh, J.L., Roozendaal, B., and
McIntyre, C.K. (2010). Memory-enhancing corticosterone treatment increases
Kim, J.J., Lee, H.J., Han, J.S., and Packard, M.G. (2001). Amygdala is critical amygdala norepinephrine and Arc protein expression in hippocampal synaptic
for stress-induced modulation of hippocampal long-term potentiation and fractions. Neurobiol. Learn. Mem. 93, 312–321.
learning. J. Neurosci. 21, 5222–5228.
Medina, A.C., Charles, J.R., Espinoza-González, V., Sánchez-Resendis, O.,
Kluen, L.M., Agorastos, A., Wiedemann, K., and Schwabe, L. (2017a). Norad- Prado-Alcalá, R.A., Roozendaal, B., and Quirarte, G.L. (2007). Glucocorticoid
renergic stimulation impairs memory generalization in women. J. Cogn. Neuro- administration into the dorsal striatum facilitates memory consolidation of
sci. 29, 1279–1291. inhibitory avoidance training but not of the context or food shock components.
Learn. Mem. 14, 673–677.
Kluen, L.M., Nixon, P., Agorastos, A., Wiedemann, K., and Schwabe, L.
(2017b). Impact of stress and glucocorticoids on schema-based learning. Meier, J.K., Staresina, B., and Schwabe, L. (2021). Stress diminishes outcome
Neuropsychopharmacology 42, 1254–1261. but enhances response representations during instrumental learning. Preprint
at bioRxiv. https://doi.org/10.1101/2021.02.12.430935.
Koolschijn, R.S., Emir, U.E., Pantelides, A.C., Nili, H., Behrens, T.E.J., and Bar-
ron, H.C. (2019). The hippocampus and neocortical inhibitory engrams protect
Morena, M., De Castro, V., Gray, J.M., Palmery, M., Trezza, V., Roozendaal, B.,
against memory interference. Neuron 101, 528–541.e6.
Hill, M.N., and Campolongo, P. (2015). Training-associated emotional arousal
Krenz, V., Sommer, T., Alink, A., Roozendaal, B., and Schwabe, L. (2021). shapes endocannabinoid modulation of spatial memory retrieval in rats.
Noradrenergic stimulation after encoding reverses the course of systems J. Neurosci. 35, 13962–13974.
consolidation in humans. Nat. Commun. 12, 6054.
Morena, M., Roozendaal, B., Trezza, V., Ratano, P., Peloso, A., Hauer, D., At-
Leal Santos, S., Stackmann, M., Muñoz Zamora, A., Mastrodonato, A., De sak, P., Trabace, L., Cuomo, V., McGaugh, J.L., et al. (2014). Endogenous
Landri, A.V., Vaughan, N., Chen, B.K., Lanio, M., and Denny, C.A.I. (2021). Pro- cannabinoid release within prefrontal-limbic pathways affects memory consol-
pranolol decreases fear expression by modulating fear memory traces. Biol. idation of emotional training. Proc. Natl. Acad. Sci. USA 111, 18333–18338.
Psychiatry 89, 1150–1161.
Moscovitch, M., and Gilboa, A. (2021). Systems consolidation, transformation
Lenow, J.K., Constantino, S.M., Daw, N.D., and Phelps, E.A. (2017). Chronic and reorganization: multiple trace theory, trace transformation theory and their
and acute stress promote overexploitation in serial decision making. competitors. Preprint at PsyArXiv. https://doi.org/10.31234/osf.io/yxbrs.
J. Neurosci. 37, 5681–5689.
Neumeister, A., Normandin, M.D., Pietrzak, R.H., Piomelli, D., Zheng, M.Q.,
Lightman, S.L., Birnie, M.T., and Conway-Campbell, B.L. (2020). Dynamics of Gujarro-Anton, A., Potenza, M.N., Bailey, C.R., Lin, S.F., Najafzadeh, S.,
ACTH and cortisol secretion and implications for disease. Endocr. Rev. 41, et al. (2013). Elevated brain cannabinoid CB1 receptor availability in post-trau-
470–490. matic stress disorder: a positron emission tomography study. Mol. Psychiatry
18, 1034–1040.
Lovitz, E.S., and Thompson, L.T. (2015). Memory-enhancing intra-basolateral
amygdala clenbuterol infusion reduces post-burst afterhyperpolarizations in Newcomer, J.W., Craft, S., Hershey, T., Askins, K., and Bardgett, M.E. (1994).
hippocampal CA1 pyramidal neurons following inhibitory avoidance learning. Glucocorticoid-induced impairment in declarative memory performance in
Neurobiol. Learn. Mem. 119, 34–41. adult humans. J. Neurosci. 14, 2047–2053.

Neuron 110, May 4, 2022 1465

 ll
Review
Nitschke, J.P., Chu, S., Pruessner, J.C., Bartz, J.A., and Sheldon, S. (2019). Reul, J.M., and de Kloet, E.R. (1985). Two receptor systems for corticosterone
Post-learning stress reduces the misinformation effect: effects of psychosocial in rat brain: microdistribution and differential occupation. Endocrinology 117,
stress on memory updating. Psychoneuroendocrinology 102, 164–171. 2505–2511.

Ohno-Shosaku, T., Maejima, T., and Kano, M. (2001). Endogenous cannabi- Roozendaal, B. (2002). Stress and memory: opposing effects of glucocorti-
noids mediate retrograde signals from depolarized postsynaptic neurons to coids on memory consolidation and memory retrieval. Neurobiol. Learn.
presynaptic terminals. Neuron 29, 729–738. Mem. 78, 578–595.

Oitzl, M.S., and de Kloet, E.R. (1992). Selective corticosteroid antagonists Roozendaal, B., Hahn, E.L., Nathan, S.V., de Quervain, D.J., and McGaugh,
modulate specific aspects of spatial orientation learning. Behav Neurosci J.L. (2004). Glucocorticoid effects on memory retrieval require concurrent
106, 62–71. noradrenergic activity in the hippocampus and basolateral amygdala.
J. Neurosci. 24, 8161–8169.
€ls, M., Schmid, W., and Cole, T.J. (1997). Spatial
Oitzl, M.S., de Kloet, E.R., Joe
learning deficits in mice with a targeted glucocorticoid receptor gene disrup- Roozendaal, B., and McGaugh, J.L. (1996). Amygdaloid nuclei lesions differ-
tion. Eur. J. Neurosci. 9, 2284–2296. entially affect glucocorticoid-induced memory enhancement in an inhibitory
avoidance task. Neurobiol. Learn. Mem. 65, 1–8.
Okuda, S., Roozendaal, B., and McGaugh, J.L. (2004). Glucocorticoid effects
on object recognition memory require training-associated emotional arousal. Roozendaal, B., and McGaugh, J.L. (2011). Memory modulation. Behav. Neu-
Proc. Natl. Acad. Sci. USA 101, 853–858. rosci. 125, 797–824.
Packard, M.G., and Goodman, J. (2012). Emotional arousal and multiple mem- Roozendaal, B., and Mirone, G. (2020). Opposite effects of noradrenergic and
ory systems in the mammalian brain. Front. Behav. Neurosci. 6, 14. glucocorticoid activation on accuracy of an episodic-like memory. Psycho-
neuroendocrinology 114, 104588.
Pape, H.C., and Paré, D. (2010). Plastic synaptic networks of the amygdala for
the acquisition, expression, and extinction of conditioned fear. Physiol. Rev. Roozendaal, B., Okuda, S., Van der Zee, E.A., and McGaugh, J.L. (2006).
90, 419–463. Glucocorticoid enhancement of memory requires arousal-induced noradren-
ergic activation in the basolateral amygdala. Proc. Natl. Acad. Sci. USA 103,
Paré, D. (2003). Role of the basolateral amygdala in memory consolidation. 6741–6746.
Prog. Neurobiol. 70, 409–420.
Roozendaal, B., Quirarte, G.L., and McGaugh, J.L. (2002). Glucocorticoids
Pavlides, C., Kimura, A., Magarinos, A.M., and McEwen, B.S. (1995). Hippo- interact with the basolateral amygdala b-adrenoceptor-cAMP/PKA system in
campal homosynaptic long-term depression/depotentiation induced by adre- influencing memory consolidation. Eur. J. Neurosci. 15, 553–560.
nal steroids. Neuroscience 68, 379–385.
Roozendaal, B., Schelling, G., and McGaugh, J.L. (2008). Corticotropin-
Peng, B., Xu, Q., Liu, J., Guo, S., Borgland, S.L., and Liu, S. (2021). Corticoste-
releasing factor in the basolateral amygdala enhances memory consolidation
rone attenuates reward-seeking behavior and increases anxiety via D2 recep-
via an interaction with the beta-adrenoceptor-cAMP pathway: dependence on
tor signaling in ventral tegmental area dopamine neurons. J. Neurosci. 41,
glucocorticoid receptor activation. J. Neurosci. 28, 6642–6651.
1566–1581.
Sandi, C., Loscertales, M., and Guaza, C. (1997). Experience-dependent facil-
Philip, N.S., Barredo, J., van ’t Wout-Frank, M., Tyrka, A.R., Price, L.H., and
itating effect of corticosterone on spatial memory formation in the water maze.
Carpenter, L.L. (2018). Network mechanisms of clinical response to transcra-
Eur. J. Neurosci. 9, 637–642.
nial magnetic stimulation in posttraumatic stress disorder and major depres-
sive disorder. Biol. Psychiatry 83, 263–272.
Sara, S.J., and Bouret, S. (2012). Orienting and reorienting: the locus coeruleus
mediates cognition through arousal. Neuron 76, 130–141.
Pitman, R.K., Rasmusson, A.M., Koenen, K.C., Shin, L.M., Orr, S.P., Gilbert-
son, M.W., Milad, M.R., and Liberzon, I. (2012). Biological studies of post-trau-
Schwabe, L., Joe€ls, M., Roozendaal, B., Wolf, O.T., and Oitzl, M.S. (2012).
matic stress disorder. Nat. Rev. Neurosci. 13, 769–787.
Stress effects on memory: an update and integration. Neurosci. Biobehav.
Popa, D., Duvarci, S., Popescu, A.T., Léna, C., and Paré, D. (2010). Coherent Rev. 36, 1740–1749.
amygdalocortical theta promotes fear memory consolidation during paradox-
ical sleep. Proc. Natl. Acad. Sci. USA 107, 6516–6519. Schwabe, L., Oitzl, M.S., Philippsen, C., Böhringer, A., Richter, S., Wippich,
€chinger, H. (2007). Stress modulates the use of spatial and stim-
W., and Scha
Popoli, M., Yan, Z., McEwen, B.S., and Sanacora, G. (2011). The stressed syn- ulus-response learning strategies in humans. Learn. Mem. 14, 109–116.
apse: the impact of stress and glucocorticoids on glutamate transmission. Nat.
Schwabe, L., Scha €chinger, H., de Kloet, E.R., and Oitzl, M.S. (2010). Cortico-
Rev. Neurosci. 13, 22–37.
steroids operate as switch between memory systems. J. Cogn. Neurosci. 22,
Preston, A.R., and Eichenbaum, H. (2013). Interplay of hippocampus and pre- 1362–1372. https://doi.org/10.1162/jocn.2009.21278.
frontal cortex in memory. Curr. Biol. 23, R764–R773.
Schwabe, L., Tegenthoff, M., Höffken, O., and Wolf, O.T. (2013). Mineralocor-
Putman, P., Hermans, E.J., Koppeschaar, H., van Schijndel, A., and van Honk, ticoid receptor blockade prevents stress-induced modulation of multiple
J. (2007). A single administration of cortisol acutely reduces preconscious memory systems in the human brain. Biol. Psychiatry 74, 801–808.
attention for fear in anxious young men. Psychoneuroendocrinology 32,
793–802. Schwabe, L., and Wolf, O.T. (2010). Learning under stress impairs memory for-
mation. Neurobiol. Learn. Mem. 93, 183–188.
Quaedflieg, C.W.E.M., Schneider, T.R., Daume, J., Engel, A.K., and Schwabe,
L. (2020). Stress impairs intentional memory control through altered theta os- Seeley, W.W. (2019). The salience network: a neural system for perceiving and
cillations in lateral parietal cortex. J. Neurosci. 40, 7739–7748. responding to homeostatic demands. J. Neurosci. 39, 9878–9882.

Quirarte, G.L., Roozendaal, B., and McGaugh, J.L. (1997). Glucocorticoid Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Kenna, H.,
enhancement of memory storage involves noradrenergic activation in the ba- Reiss, A.L., and Greicius, M.D. (2007). Dissociable intrinsic connectivity net-
solateral amygdala. Proc. Natl. Acad. Sci. USA 94, 14048–14053. works for salience processing and executive control. J. Neurosci. 27,
2349–2356.
Raichle, M.E. (2013). The Brain’s default mode network. Annu. Rev. Neurosci.
38, 1–15. Seidenbecher, T., Laxmi, T.R., Stork, O., and Pape, H.-C. (2003). Amygdalar
and hippocampal theta rhythm synchronization During fear memory retrieval.
Raio, C.M., Hartley, C.A., Orederu, T.A., Li, J., and Phelps, E.A. (2017). Stress Science 301, 846–850.
attenuates the flexible updating of aversive value. Proc. Natl. Acad. Sci. USA
114, 11241–11246. Seo, D.O., Zhang, E.T., Piantadosi, S.C., Marcus, D.J., Motard, L.E., Kan, B.K.,
Gomez, A.M., Nguyen, T.K., Xia, L., and Bruchas, M.R. (2021). A locus coeru-
Ranganath, C., and Ritchey, M. (2012). Two cortical systems for memory- leus to dentate gyrus noradrenergic circuit modulates aversive contextual pro-
guided behaviour. Nat. Rev. Neurosci. 13, 713–726. cessing. Neuron 109, 2116–2130.e6.

1466 Neuron 110, May 4, 2022

 ll
Review
Shi, M.M., Fan, K.M., Qiao, Y.N., Xu, J.H., Qiu, L.J., Li, X., Liu, Y., Qian, Z.Q., Van Stegeren, A.H., Roozendaal, B., Kindt, M., Wolf, O.T., and Joe €ls, M.
Wei, C.L., Han, J., et al. (2020). Hippocampal m-opiond receptors on (2010). Interacting noradrenergic and corticosteroid systems shift human brain
GABAergic neurons mediate stress-induced impairment of memory retrieval. activation patterns during encoding. Neurobiol. Learn. Mem. 93, 56–65.
Mol. Psychiatry 25, 977–992.
Vandael, D., Wierda, K., Vints, K., Baatsen, P., De Groef, L., Moons, L., Ryba-
Shields, G.S., Sazma, M.A., and Yonelinas, A.P. (2016). The effects of acute kin, V., and Gounko, N.V. (2021). Corticotropin-releasing factor induces func-
stress on core executive functions: a meta-analysis and comparison with tional and structural synaptic remodelling in acute stress. Transl. Psychiatry
cortisol. Neurosci. Biobehav. Rev. 68, 651–668. 11, 378.

Shin, J.D., Tang, W., and Jadhav, S.P. (2019). Dynamics of awake hippocam- Vincent, J.L., Kahn, I., Snyder, A.Z., Raichle, M.E., and Buckner, R.L. (2008).
pal-prefrontal replay for spatial learning and memory-guided decision making. Evidence for a frontoparietal control system revealed by intrinsic functional
Neuron 104, 1110–1125.e7. connectivity. J. Neurophysiol. 100, 3328–3342.

Visser, R.M., Henson, R.N., and Holmes, E.A. (2021). A naturalistic paradigm to
Shohamy, D., and Adcock, R.A. (2010). Dopamine and adaptive memory.
investigate post-encoding neural activation patterns in relation to subsequent
Trends Cogn. Sci. 14, 464–472.
voluntary and intrusive recall of distressing events. Biol. Psychiatry Cogn. Neu-
rosci. Neuroimaging. https://doi.org/10.1016/j.bpsc.2021.08.006.
Shrestha, P., Shan, Z., Mamcarz, M., Ruiz, K.S.A., Zerihoun, A.T., Juan, C.Y.,
Herrero-Vidal, P.M., Pelletier, J., Heintz, N., and Klann, E. (2020). Amygdala €ls, M., and Schwabe, L. (2016). Cognitive adap-
Vogel, S., Fernández, G., Joe
inhibitory neurons as loci for translation in emotional memories. Nature 586, tation under stress: a case for the mineralocorticoid receptor. Trends Cogn.
407–411. Sci. 20, 192–203.
Siller-Pérez, C., Fuentes-Ibañez, A., Sotelo-Barrera, E.L., Serafı́n, N., Prado- Vogel, S., Kluen, L.M., Fernández, G., and Schwabe, L. (2018). Stress affects
Alcalá, R.A., Campolongo, P., Roozendaal, B., and Quirarte, G.L. (2019). the neural ensemble underlying the integration of new information and prior
Glucocorticoid interactions with the dorsal striatal endocannabinoid system knowledge. Neuroimage 173, 176–187.
in regulating inhibitory avoidance memory. Psychoneuroendocrinology
99, 97–103. Vogel, S., Klumpers, F., Schröder, T.N., Oplaat, K.T., Krugers, H.J., Oitzl, M.S.,
€ls, M., Doeller, C.F., and Fernández, G. (2017). Stress induces a shift to-
Joe
Siller-Pérez, C., Serafı́n, N., Prado-Alcalá, R.A., Roozendaal, B., and Quirarte, wards striatum-dependent stimulus-response learning via the mineralocorti-
G.L. (2017). Glucocorticoid administration into the dorsolateral but not dorso- coid receptor. Neuropsychopharmacology 42, 1262–1271.
medial striatum accelerates the shift from a spatial toward procedural memory.
Neurobiol. Learn. Mem. 141, 124–133. Vogel, S., and Schwabe, L. (2016). Stress in the zoo: tracking the impact of
stress on memory formation over time. Psychoneuroendocrinology 71, 64–72.
Simon-Kutscher, K., Wanke, N., Hiller, C., and Schwabe, L. (2019). Fear
without context: acute stress modulates the balance of cue-dependent and Vouimba, R.M., Yaniv, D., and Richter-Levin, G. (2007). Corticotropin-
contextual fear learning. Psychol. Sci. 30, 1123–1135. releasing factor induces functional and structural synaptic remodelling in acute
stress. Neuropharmacology 52, 244–252.
Singer, A.C., and Frank, L.M. (2009). Rewarded outcomes enhance reactiva-
tion of experience in the hippocampus. Neuron 64, 910–921. Vukojevic, V., Coynel, D., Ghaffari, N.R., Freytag, V., Elbert, T., Kolassa, I.T.,
Wilker, S., McGaugh, J.L., Papassotiropoulos, A., and De Quervain, D.J.
(2020). NTRK2 methylation is related to reduced PTSD risk in two African co-
Soravia, L.M., Moggi, F., and De Quervain, D.J. (2021). Effects of cortisol
administration on craving during in vivo exposure in patients with alcohol horts of trauma survivors. Proc. Natl. Acad. Sci. USA 117, 21667–21672.
use disorder. Transl. Psychiatry 11, 6. Wiegert, O., Joe €ls, M., and Krugers, H.J. (2006). Timing is essential for rapid
effects of corticosterone on synaptic potentiation in the mouse hippocampus.
Squire, L.R. (1992). Memory and hippocampus: a synthesis from findings with Learn. Mem. 13, 110–113.
rats, monkeys and humans. Psychol Rev 99, 195–231.
Wirz, L., Bogdanov, M., and Schwabe, L. (2018). Habits under stress: mecha-
Squire, L.R., and Alvarez, P. (1995). Retrograde amnesia and memory consol- nistic insights across different types of learning. Curr. Opin. Behav. Sci. 20, 9–16.
idation: a neurobiological perspective. Curr. Opin. Neurobiol. 5, 169–177.
Wirz, L., Reuter, M., Wacker, J., Felten, A., and Schwabe, L. (2017a). A haplotype
Takehara-Nishiuchi, K., and McNaughton, B.L. (2008). Spontaneous changes associated with enhanced mineralocorticoid receptor expression facilitates the
of neocortical code for associative memory during consolidation. Science 322, stress-induced shift from ‘cognitive’ to ‘habit’ learning. eNeuro 4, 1–16.
960–963.
Wirz, L., Wacker, J., Felten, A., Reuter, M., and Schwabe, L. (2017b). A deletion
Tambini, A., and Davachi, L. (2019). Awake reactivation of prior experiences variant of the a2b-adrenoceptor modulates the stress-induced shift from
consolidates memories and biases cognition. Trends Cogn. Sci. 23, 876–890. ‘cognitive’ to ‘habit’ memory. J. Neurosci. 37, 2149–2160.

Tertil, M., Skupio, U., Barut, J., Dubovyk, V., Wawrzczak-Bargiela, A., Soltys, Wu, C.-T., Haggerty, D., Kemere, C., and Ji, D. (2017). Hippocampal awake
Z., Golda, S., Kudla, L., Wiktorowska, L., Szklarczyk, K., et al. (2018). Gluco- replay in fear memory retrieval. Nat. Neurosci. 20, 571–580.
corticoid receptor signaling in astrocytes is required for aversive memory for-
mation. Transl. Psychiatry 8, 255. Young, C.B., Raz, G., Everaerd, D., Beckmann, C.F., Tendolkar, I., Hendler, T.,
Fernández, G., and Hermans, E.J. (2017). Dynamic shifts in large-scale brain
Valentino, R.J., and van Bockstaele, E.J. (2005). Functional interactions be- network balance as a function of arousal. J. Neurosci. 37, 281–290.
tween stress neuromediators and the locus coeruleus–norepinephrine system.
Yuen, E.Y., Liu, W., Karatsoreos, I.N., Feng, J., McEwen, B.S., and Yan, Z.
Tech. Behav. Neurosci. 15, 465–486.
(2009). Acute stress enhances glutamatergic transmission in prefrontal cortex
and facilitates working memory. Proc. Natl. Acad. Sci. USA 106, 14075–14079.
€ls, M., and Kindt, M. (2013). Time-
van Ast, V.A., Cornelisse, S., Meeter, M., Joe
dependent effects of cortisol on the contextualization of emotional memories. Zerbes, G., Kausche, F.M., and Schwabe, L. (2020). Stress-induced cortisol
Biol. Psychiatry 74, 809–816. modulates the control of memory retrieval towards the dorsal striatum. Eur.
J. Neurosci. https://doi.org/10.1111/ejn.14942.
Van Kesteren, M.T.R., Fernández, G., Norris, D.G., and Hermans, E.J. (2010).
Persistent schema-dependent hippocampal-neocortical connectivity during Zerbes, G., and Schwabe, L. (2021). Stress-induced bias of multiple memory
memory encoding and postencoding rest in humans. Proc. Natl. Acad. Sci. systems during retrieval depends on training intensity. Psychoneuroendocri-
USA 107, 7550–7555. nology 130, 105281.

van Leeuwen, J.M.C., Vink, M., Joe €ls, M., Kahn, R.S., Hermans, E.J., and Zerbi, V., Floriou-Servou, A., Markicevic, M., Vermeiren, Y., Sturman, O., Priv-
Vinkers, C.H. (2019). Reward-related striatal responses following stress in itera, M., von Ziegler, L., Ferrari, K.D., Weber, B., de Deyn, P.P., et al. (2019).
healthy individuals and patients With bipolar disorder. Biol. Psychiatry Cogn. Rapid reconfiguration of the functional connectome after chemogenetic locus
Neurosci. Neuroimaging 4, 966–974. coeruleus activation. Neuron 103, 702–718.e5.

Neuron 110, May 4, 2022 1467