Adaptive memory systems for remembering the salient and the seemingly mundane

Exercise and Fear and Safety Learning

Fear conditioning paradigms have been studied for over 100 years and are of great interest to the behavioral and clinical sciences given that several safety learning processes (e.g., extinction learning and recall) are thought to be fundamental to the success of exposure-based therapies for anxiety and related disorders. This chapter provides an overview of preclinical and clinical investigations that examined the effects of exercise on initial fear acquisition, fear extinction learning and consolidation, and return of fear outcomes. This chapter highlights the collective body of evidence suggesting that exercise administered after extinction learning enhances the consolidation and subsequent recall of extinction memories to a greater extent than exercise administered prior to extinction learning. This suggests that the addition of exercise after exposure therapy sessions may improve treatment outcomes for people with anxiety and related disorders. Potential mechanisms are discussed in addition to suggestions for future research to improve our understanding of the effects of exercise on fear conditioning and extinction outcomes.

Is there selective retroactive memory enhancement in humans?: a meta-analysis

Memory is an adaptive and flexible system that preferentially stores motivationally relevant information. However, in some cases information that is initially irrelevant can become relevant at a later time. The question arises whether and to what extent the memory system can retroactively boost memories of the initially irrelevant information. Experimental studies in animals and humans have provided evidence for such retroactive memory boosting. Additionally, these studies suggest that retroactive memory enhancement (RME) can be selective to the semantic meaning of the material. Nonetheless, recent experimental work could not replicate these findings, posing the question whether the selective RME effect is reliable. To synthesize the available evidence, we conducted meta-analyses of 14 experiments. Although the classical meta-analytic procedure suggested a small selective RME effect, Cohen's dz = 0.16, when accounting for small-study bias using robust Bayesian meta-analysis the null hypothesis was supported, Cohen's dz = 0.02, BF01 = 3.03. Furthermore strong evidence was found for a bias due to small-study effects, BF10 = 11.39. Together, this calls the reliability of a selective RME effect into question.

Emotional learning retroactively promotes memory integration through rapid neural reactivation and reorganization

Neutral events preceding emotional experiences can be better remembered, likely by assigning them as significant to guide possible use in future. Yet, the neurobiological mechanisms of how emotional learning enhances memory for past mundane events remain unclear. By two behavioral studies and one functional magnetic resonance imaging study with an adapted sensory preconditioning paradigm, we show rapid neural reactivation and connectivity changes underlying emotion-charged retroactive memory enhancement. Behaviorally, emotional learning retroactively enhanced initial memory for neutral associations across the three studies. Neurally, emotional learning potentiated trial-specific reactivation of overlapping neural traces in the hippocampus and stimulus-relevant neocortex. It further induced rapid hippocampal-neocortical functional reorganization supporting such retroactive memory benefit, as characterized by enhanced hippocampal-neocortical coupling modulated by the amygdala during emotional learning, and a shift of hippocampal connectivity from stimulus-relevant neocortex to distributed transmodal prefrontal-parietal areas at post-learning rests. Together, emotional learning retroactively promotes memory integration for past neutral events through stimulating trial-specific reactivation of overlapping representations and reorganization of associated memories into an integrated network to foster its priority for future use.

Survival of the salient: Aversive learning rescues otherwise forgettable memories via neural reactivation and post-encoding hippocampal connectivity

The effects of aversive events on memory are complex and go beyond the simple enhancement of threatening information. Negative experiences can also rescue related but otherwise forgettable details encoded close in time. Here, we used functional magnetic resonance imaging (fMRI) in healthy young adults to examine the brain mechanisms that support this retrograde memory effect. In a two-phase incidental encoding paradigm, participants viewed different pictures of tools and animals before and during Pavlovian fear conditioning. During Phase 1, these images were intermixed with neutral scenes, which provided a unique 'context tag' for this specific phase of encoding. A few minutes later, during Phase 2, new pictures from one category were paired with a mild shock (threat-conditioned stimulus; CS+), while pictures from the other category were not shocked. FMRI analyses revealed that, across-participants, individuals who showed aversive learning-related retroactive memory benefits for Phase 1 CS+ items were also more likely to exhibit three brain effects: first, greater spontaneous reinstatement of the Phase 1 context when participants viewed conceptually-related CS+ items in Phase 2; second, greater successful encoding-related VTA/SN and LC activation for Phase 2 CS+ items; and third, learning-dependent increases in post-encoding hippocampal functional coupling with CS+ category-selective cortex. These biases in hippocampal-cortical connectivity also mediated the relationship between VTA/SN aversive encoding effects and across-participant variability in the retroactive memory benefit. Collectively, our findings suggest that both online and offline brain mechanisms may enable threatening events to preserve memories that acquire new significance in the future.

Retroactive and graded prioritization of memory by reward

Many decisions are based on an internal model of the world. Yet, how such a model is constructed from experience and represented in memory remains unknown. We test the hypothesis that reward shapes memory for sequences of events by retroactively prioritizing memory for objects as a function of their distance from reward. Human participants encountered neutral objects while exploring a series of mazes for reward. Across six data sets, we find that reward systematically modulates memory for neutral objects, retroactively prioritizing memory for objects closest to the reward. This effect of reward on memory emerges only after a 24-hour delay and is stronger for mazes followed by a longer rest interval, suggesting a role for post-reward replay and overnight consolidation, as predicted by neurobiological data in animals. These findings demonstrate that reward retroactively prioritizes memory along a sequential gradient, consistent with the role of memory in supporting adaptive decision-making.