Skip to main content

Deciphering neural codes of valence-based emotional memories

Final Report Summary - NEURALCODES_EMO (Deciphering neural codes of valence-based emotional memories)

Understanding emotional memories at a detailed level in a primate model is a crucial step in our goal to understand ourselves, and a required step to allow specific therapeutic interventions for a spectrum of psychiatric disorders. Although major advances were recently made in rodent models, the circuitry in primates is highly evolved with complex inter-regional interactions. The current project was designed to develop new models for emotional learning in primates, new models for flexibility of emotional memories, and thus allow detailed examination of the underlying neural mechanisms by recording the simultaneous activity of neurons in amygdala-prefrontal regions and understanding their role in normal and abnormal emotional manifestations - with a focus on anxiety.
We were able to develop several behavioral models for emotional learning, and then use these models to further develop models of flexibility of learning. We focused on two types of flexibility: safety learning and over-generalization. Failure of safety learning and/or over-generalization of fear memories are used to probe mechanisms of learning in the primate brain and serve as models for anxiety-disorders. We identify neural interactions between the amygdala and the prefrontal cortex that underlie learning signals as predicted by learning theory, and show that these interactions can underlie successful and maladaptive states of learning and memory under these emotional conditions. By combining approaches from cognitive psychology, learning theory, and large-scale electrophysiology in behaving primates, we shed important light on mechanisms of normal emotional learning and design models for neuropsychiatric conditions.