A new study reports that a behavioral intervention can do more than suppress the outward expression of fear: it may erase the neural trace of a traumatic memory from a defined population of cells in the brain. In research published in Experimental & Molecular Medicine, Li, Yang, Cui and colleagues describe how an “extinction-restraint” intervention eliminated fear memory encoded by engram cells in the basolateral amygdala, or BLA. The effect depended on dopamine signaling through D2 receptors and on a previously underappreciated source of that dopamine, the locus coeruleus. The findings offer a mechanistic explanation for why some fear memories remain resistant to conventional extinction training—and point toward new strategies for treating persistent anxiety and trauma-related disorders.
Fear memories are not stored as diffuse impressions spread evenly across the brain. Instead, learning recruits ensembles of neurons that become especially responsive when the memory is later retrieved. These ensembles, often called engram cells, can be identified by their activity during memory formation and reactivation. The BLA is a crucial component of this system. It integrates sensory information with emotional significance and helps determine whether a stimulus should trigger defensive responses. When an otherwise neutral cue becomes associated with danger, activity in BLA circuits can rapidly mobilize fear-related behavior, changes in heart rate and stress hormones, and heightened vigilance.
Extinction training is commonly used to weaken these responses. During extinction, an organism repeatedly encounters the formerly threatening cue without the aversive event that originally accompanied it. This process usually creates a new inhibitory memory rather than deleting the original one. The brain learns that the cue is now safe, but the old fear association can return under stress, in a different environment, or after the passage of time. That relapse problem is one reason exposure-based therapies, although highly effective for many people, do not always produce permanent relief. The new study addresses this central limitation by asking whether a carefully designed intervention can act directly on the fear engram rather than simply placing a competing safety memory on top of it.
The researchers focused on communication between the BLA and the locus coeruleus, a small brainstem nucleus best known for producing noradrenaline and regulating arousal, attention and responses to novelty. The locus coeruleus is increasingly understood as a flexible neuromodulatory hub rather than a simple “stress center.” Under particular conditions, its neurons can also provide dopamine to forebrain regions. Dopamine changes the way neural circuits respond to incoming information by influencing excitability, synaptic plasticity and the assignment of motivational value. In the BLA, these effects may determine whether an existing fear association is preserved, updated or weakened.
The study identifies dopamine D2 receptors as a critical molecular switch in this process. D2 receptors belong to the family of G-protein-coupled dopamine receptors and generally reduce cellular signaling through inhibitory intracellular pathways, although their effects depend on the cell type and circuit in which they are expressed. By examining the intervention’s effects on BLA engram cells and manipulating D2-related signaling, the researchers found that the fear-memory-erasing outcome required this receptor pathway. When D2 signaling was disrupted, the intervention no longer produced the same elimination of the engram-associated fear response, indicating that dopamine was not merely correlated with the behavioral change but functionally involved in it.
The result is particularly striking because it challenges a simplified view of memory erasure. A fear engram is not a permanent physical inscription in a single neuron, nor is it necessarily destroyed by one molecular event. Memories are distributed across ensembles whose connections and activity patterns change over time. The researchers’ findings suggest that extinction-restraint may push the relevant BLA ensemble into a state in which its original fear-related representation is no longer retrievable or functionally dominant. In other words, the intervention may alter the circuit’s ability to reactivate the old pattern, rather than simply silencing the cells temporarily. That distinction matters because temporary suppression can disappear when the organism encounters a new threat or a stressful context.
The locus coeruleus appears to provide the neuromodulatory signal that makes this remodeling possible. By releasing dopamine into the BLA during the intervention, locus-coeruleus neurons may mark the experience as sufficiently important to trigger synaptic reorganization. D2-receptor activation could then modify the strength of connections among BLA neurons, alter the balance between excitatory and inhibitory inputs, or change how the fear ensemble responds to the extinguished cue. The study’s model places the brainstem nucleus in direct control of a memory-updating process in the amygdala, revealing a circuit route through which arousal-related systems can influence whether an emotional memory persists.
The term “extinction-restraint” describes the distinctive behavioral framework examined by the researchers. Rather than relying on ordinary extinction exposure alone, the protocol appears to combine extinction learning with a controlled restraint component, creating a state in which the animal must process the previously threatening cue under tightly regulated conditions. Such an approach may engage attention, arousal and learning systems more strongly than standard extinction, allowing the safety information to compete with—or overwrite—the original fear representation. The precise behavioral parameters will be important for interpreting the work and for determining whether the effect depends on restraint itself, on the timing of the intervention, or on the interaction between controlled stress and extinction learning.
The implications extend beyond one fear-conditioning paradigm. If the findings can be replicated and translated, treatments might eventually be designed to recruit the same locus-coeruleus-to-BLA dopamine pathway during exposure therapy. Pharmacological agents that selectively influence D2 signaling, stimulation of neuromodulatory circuits, or precisely timed behavioral interventions could potentially make extinction memories more durable. However, the distance between an experimental memory model and a clinical treatment remains substantial. Fear learning in laboratory animals is highly controlled, while human traumatic memories are often complex, repeatedly reinforced and intertwined with language, identity and social experience. Manipulating dopamine receptors also carries risks, because dopamine systems participate in movement, motivation, reward learning and psychosis-related processes.
The work therefore offers a powerful biological proof of principle rather than an immediate cure for post-traumatic stress disorder or phobias. Its most important contribution may be the identification of a specific circuit and receptor mechanism that distinguishes memory suppression from memory revision. By showing that locus-coeruleus-derived dopamine can act through D2 receptors in BLA fear engram cells, the study provides a molecular explanation for how an intervention can produce a deeper change in emotional memory. The next questions will be whether the erased representation remains absent after long delays, whether fear can return in unfamiliar contexts, how the intervention affects other memories, and whether comparable circuitry operates in humans. For now, the findings suggest that the brain’s most stubborn fear memories may be more biologically negotiable than previously believed.
Subject of Research: Fear-memory erasure, basolateral amygdala engram cells, dopamine D2 receptor signaling, and locus-coeruleus modulation during extinction-restraint intervention
Article Title: Extinction-restraint intervention eradicates fear memory encoded in BLA engram cells via D2 receptor-mediated dopaminergic modulation from locus coeruleus
Article References: Li, T., Yang, J., Cui, C. et al. “Extinction-restraint intervention eradicates fear memory encoded in BLA engram cells via D2 receptor-mediated dopaminergic modulation from locus coeruleus.” Experimental & Molecular Medicine (2026). https://doi.org/10.1038/s12276-026-01794-0
Image Credits: AI Generated
DOI: 10.1038/s12276-026-01794-0
Keywords: fear memory, extinction learning, restraint intervention, engram cells, basolateral amygdala, locus coeruleus, dopamine, D2 receptor, emotional memory, neuroscience
Tags: brain circuits involved in fear learningdopamine D2 receptor signalingengram cells in basolateral amygdalaextinction-restraint interventionfear memory erasurelocus coeruleus dopamine sourceneural basis of fear memory resistanceneural mechanisms of fear extinctionpersistent anxiety disorder treatmenttargeted neural interventions for fear extinctiontrauma-related disorder therapytraumatic memory elimination
