In their superb glutamate amplifies noradrenergic effects (GANE) model, Mather and colleagues convincingly argue that under conditions of arousal-induced phasic activity of the locus coeruleus (LC), locally elevated glutamate (GLU) levels amplify noradrenergic (norepinephrine [NE]) release from the LC, thus creating functional hotspots of prioritized processing that bias perception and memory. Although the GANE model focuses on stimulus salience coding through rapid GLU and NE signaling and their focal interactions, it should be emphasized that endocrine signals, including the adrenal stress hormone cortisol (CORT), brain concentrations of which peak within minutes as a result of hypothalamus–pituitary–adrenal (HPA) axis activation (de Kloet et al. Reference de Kloet, Joels and Holsboer2005), also intimately interact with NE to code perceptual and mnemonic priority, especially under conditions of emotional arousal.

In functional magnetic resonance imaging (fMRI) experiments, emotional arousal is frequently operationalized by exposing subjects to facial displays of emotion, which evoke responses in specific functional subdivisions of the amygdala (Goossens et al. Reference Goossens, Kukolja, Onur, Fink, Maier, Griez, Schruers and Hurlemann2009; Hurlemann et al. Reference Hurlemann, Rehme, Diessel, Kukolja, Maier, Walter and Cohen2008). One established means of segregating the neuromodulatory effects produced by NE, CORT, and their interactions, is pharmacologic fMRI (phMRI) (Patin & Hurlemann Reference Patin and Hurlemann2011). A combination of phMRI with histoprobabilistic maps of the subregional architecture of the amygdala (Goossens et al. Reference Goossens, Kukolja, Onur, Fink, Maier, Griez, Schruers and Hurlemann2009; Hurlemann et al. Reference Hurlemann, Rehme, Diessel, Kukolja, Maier, Walter and Cohen2008) revealed that blockade of β-noradrenergic receptors with the non-specific antagonist propranolol (40 mg po) desensitized the basolateral amygdala (BLA) (Hurlemann et al. Reference Hurlemann, Walter, Rehme, Kukolja, Santoro, Schmidt, Schnell, Musshoff, Keysers, Maier, Kendrick and Onur2010), which is consistent with behavioral data indicating that propranolol (40 mg po) eliminated a facilitation of declarative learning from facial feedback (Mihov et al. Reference Mihov, Mayer, Musshoff, Maier, Kendrick and Hurlemann2010). In contrast, enhancement of BLA reactivity with the NE re-uptake inhibitor (NARI) reboxetine (4 mg po) produced a response bias toward fearful faces (Onur et al. Reference Onur, Walter, Schlaepfer, Rehme, Schmidt, Keysers, Maier and Hurlemann2009). Together, these results suggest that increases in NE signaling may be essential for converting the BLA – an area of the brain controlled by powerful inhibitory circuits (Ehrlich et al. Reference Ehrlich, Humeau, Grenier, Ciocchi, Herry and Luthi2009) – into a fear module (Onur et al. Reference Onur, Walter, Schlaepfer, Rehme, Schmidt, Keysers, Maier and Hurlemann2009). One interpretation of these findings is that phasic increases in endogenous NE signaling per se might be sufficient to code stimulus salience. However, because of its pivotal role in orchestrating fear memory acquisition and storage via N-methyl-D-aspartate (NMDA) receptor-mediated long-term potentiation (LTP) (Ehrlich et al. Reference Ehrlich, Humeau, Grenier, Ciocchi, Herry and Luthi2009), the BLA may be a locus of extensive GLU–NE interactions, such that observations of a reboxetine-induced increase in BLA signals may, in fact, support the GANE model.

In addition to rapid neuromodulatory effects mediated by NE per se, emotional arousal elicits heightened adrenal release of CORT, which feeds back on the amygdala and hippocampus via activation of mineralocorticoid and glucocorticoid receptors in these regions (de Kloet et al. Reference de Kloet, Joels and Holsboer2005; McEwen et al. Reference McEwen, Bowles, Gray, Hill, Hunter, Karatsoreos and Nasca2015). Experimentally, this endocrine response can be mimicked by exogenous administration of synthetic CORT (20–40 mg po), and studies based on this challenge not only have noted a desensitization of the amygdala during fear conditioning (Merz et al. Reference Merz, Tabbert, Schweckendiek, Klucken, Vaitl, Stark and Wolf2010) and reward anticipation (Montoya et al. Reference Montoya, Bos, Terburg, Rosenberger and van Honk2014), but also have detected timing-dependent changes in hippocampal memory functions. Specifically, when coinciding with declarative memory encoding, stress levels of CORT enhance long-term recall (Buchanan & Lovallo Reference Buchanan and Lovallo2001), whereas their occurrence during retrieval impairs performance (de Quervain et al. Reference de Quervain, Roozendaal, Nitsch, McGaugh and Hock2000).

Most important, endogenous CORT and NE signals do not act in isolation, and there is accumulating experimental evidence that coactivation of both systems under emotional arousal is crucial for facilitating amygdala–hippocampus interplay during declarative memory formation. The resultant advantage of privileged declarative encoding of salient stimuli, however, comes at the expense of reduced recall of preceding and following information. This peri-emotional amnesia is BLA as well as β-noradrenergic dependent (Hurlemann Reference Hurlemann2006; Hurlemann et al. Reference Hurlemann, Hawellek, Matusch, Kolsch, Wollersen, Madea, Vogeley, Maier and Dolan2005; Reference Hurlemann, Matusch, Hawellek, Klingmuller, Kolsch, Maier and Dolan2007a; Reference Hurlemann, Wagner, Hawellek, Reich, Pieperhoff, Amunts, Oros-Peusquens, Shah, Maier and Dolan2007b; Strange et al. Reference Strange, Hurlemann and Dolan2003) and further amplified, in both magnitude and temporal extent, by combined prelearning administration of exogenous CORT (30 mg po) and reboxetine (4 mg po), thus suggesting synergistic NE–CORT interactions (Hurlemann Reference Hurlemann2008; Hurlemann et al. Reference Hurlemann, Matusch, Hawellek, Klingmuller, Kolsch, Maier and Dolan2007a). The same pharmacologic intervention was found to induce a negative response bias toward fearful faces in the centromedial nucleus of the amygdala (CMA), an effect that was absent when CORT levels were augmented alone (Kukolja et al. Reference Kukolja, Schlapfer, Keysers, Klingmuller, Maier, Fink and Hurlemann2008). Evidence indicates that response shifts mediated by CORT, NE, and their interactions are not restricted to the CMA, but propagate to interconnected areas including the dorsal striatum, which can be prevented by blockade of mineralocorticoid receptors with spironolactone (400 mg po) (Vogel et al. Reference Vogel, Klumpers, Krugers, Fang, Oplaat, Oitzl, Joels and Fernandez2015b).

Collectively, these findings argue for a reallocation of neural resources as a function of CORT and NE coactivation under emotional arousal, hence enabling prioritized access to the salience network and memory stores. Obviously, this mechanism confers costs and benefits, evident in a larger devotion of amygdala–hippocampal resources during encoding (Kukolja et al. Reference Kukolja, Klingmuller, Maier, Fink and Hurlemann2011) and deactivation of prefrontal cortex (PFC) (van Stegeren et al. Reference van Stegeren, Roozendaal, Kindt, Wolf and Joels2010). It has been conceptualized that such co-occurrence of deficient top-down control from PFC and enhanced amygdala–hippocampus interactions under conditions of heightened CORT and NE release may result in hypermnesia for emotional events, which, when manifest in extreme forms, is pathognomonic of post-traumatic stress disorder (PTSD) (Hurlemann Reference Hurlemann2008). Converging support for this etiologic model comes from preclinical (Bryant et al. Reference Bryant, McGrath and Felmingham2013) and clinical studies (Nicholson et al. Reference Nicholson, Bryant and Felmingham2014), both of which suggest that NE and CORT co-activation predisposes to the development of indelible memories. Future research addressing the mechanistic underpinnings of arousal-induced memory distortions in PTSD should, therefore, not only focus on neurotransmitter interactions between GLU and NE, as outlined by the GANE model, but also take the interplay of NE and endocrine players including CORT into perspective, which promotes stress-induced remodeling of neural architecture through (epi)genetic modifications as well as rapid non-genomic adaptations (de Kloet et al. Reference de Kloet, Joels and Holsboer2005; McEwen et al. Reference McEwen, Bowles, Gray, Hill, Hunter, Karatsoreos and Nasca2015). The latter include the non-genomic modulation of hippocampal glutamate transmission via activation of mineralocorticoid receptors (Karst et al. Reference Karst, Berger, Turiault, Tronche, Schütz and Joëls2005), further illustrating the rapid susceptibility of memory functions to emotional arousal and stress.