Distinct medial-temporal lobe mechanisms of encoding and amygdala-mediated memory reinstatement for disgust and fear

Abstract

Many associative memory traces are charged with emotion that can either impair or enhance subsequent retrieval. Various factors may influence such emotion effects, including the nature of associations and type of emotions. We show that long-term recognition memory of very close verbal associations was enhanced for emotionally charged material, and these effects differed between two negative emotion categories matched for arousal level. Specifically, memory was better for word pairs related to disgust than word pairs related to fear. Furthermore, these two emotions distinctively modulated neural processes during encoding and its relationship with retrieval, regardless of arousal evoked by word pairs. Amygdala and perirhinal cortex activity were associated with enhanced memory for disgust-related unitizations, whereas better memory for fear-related unitizations engaged parahippocampal cortex and hippocampus. Finally, the magnitude of amygdala activation during encoding was related to the subsequent fidelity of hippocampal activity patterns during retrieval. These results show that dissociable neural pathways and reinstatement mechanisms are involved in associative memory for different emotion categories and that arousal alone cannot fully explain emotional influences on associative memory.individually control for the effects of valence and arousal dimensions. Most importantly, this paradigm enabled us to compare the encoding-retrieval similarity of brain activation patterns elicited across two different emotion categories and determine which brain regions promoted this measure of neural reinstatement.We found a memory enhancement effect due to emotion. Moreover, word pairs related to disgust were remembered better than word pairs related to fear, and this effect could not be explained by differences in arousal. We found distinctive neural pathways preferentially engaged in encoding modulation by disgust or fear. Specifically, the left AMY and left PRC were more activated during successful encoding of disgust-related unitized word pairs, whereas right PHC and HC were more activated for fear-related unitized word pairs, even when the level of arousal was regressed out from the analysis. Moreover, during successful encoding of disgust-compared to fearrelated word pairs, we observed an increase in functional connectivity between AMY, PRC and PHC. Finally, for disgusting compared to fearful word pairs, the left AMY encoding activation was found to be critical for the level of trial-by-trial encodingretrieval similarity of brain activation patterns.Altogether, the results show that behavioural and neuronal mechanisms of long-term associative memory can be differently modulated by basic emotion categories, over and above of valence and arousal dimensions. For the first time, we revealed that due to AMY and PRC engagement, disgust as opposed to fear has a strong influence on reinstatement and recognition memory of unitized word pairs, which is in line with its distinctive and perhaps evolutionary-driven role in memorizing ...

Emotion schema effects on associative memory differ across emotion categories at the behavioural, physiological and neural level

“…Our recent study showed distinct engagement of key brain regions supporting emotional memory, i.e. amygdala (AMY) and MTL areas during encoding and reinstatement of disgust and fear (Riegel et al, 2022). However, it remains unknown how memory modulation by disgust and fear is related to mechanisms typical to schemas such as congruency and incongruency effects, as well as what are the concomitant physiological and neural mechanisms.…”

“…Labeling facial expressions related to disgust and fear, inherent to our experimental task, was shown to enhance encoding of contextual information (Barrett & Kensinger, 2010), yet differences between them were not explored. Our recent findings showed that these two emotions could imply different sensory and semantic processing of events and their context (Riegel et al, 2022), suggesting that they could differentially modulate congruency and incongruency effects. Therefore, memory encoding might disclose not only emotion-specific effects of the event content, but also additive or interactive effects of an emotional congruency with the context, both described above.…”

“…Likewise, we would expect to see the emotion- specific effects during encoding at the level of visual attention - how often and for how long we look at words vs, faces, for different predictions see: (Fink-Lamotte, Svensson, Schmitz, & Exner, 2021b; Schienle et al, 2021; Schienle, Übel, Gremsl, Schöngassner, & Körner, 2016); and physiological reactions - pupil response reflecting arousal (Aston-Jones & Cohen, 2005b), pupil temporal dynamics reflecting sympathetic vs. parasympathetic activation (Clewett, Gasser, & Davachi, 2020a; Johansson, Pärnamets, Bjernestedt, & Johansson, 2018a). Moreover, we would expect to observe different neural mechanism of encoding for disgust- vs. fear-related pairs (distinct engagement of AMY and MTL regions (Riegel et al, 2022).…”