Dissection and manipulation of fear memory trace with artificial control of neural activity

Abstract

Memory is thought to be sparsely encoded throughout multiple brain regions forming unique memory trace and undergo constant change during the internal and external reactivation of those brain regions. Adopting classical auditory fear conditioning as a behavior model, I conducted three independent researches exploring the spatial and temporal dynamics of fear memory. In the first experiment, brain-region specific activity in recent and remote auditory fear memory retrieval was tested. Although evidences have established that the amygdala is a key brain site for memory storage and retrieval of auditory conditioned fear memory, the learning-related changes are also found in upstream sensory regions such as auditory thalamus and cortex. To investigate how the auditory upstream takes part in fear memory process at different time phase, I systematically imaged the brain activity patterns in the lateral amygdala (LA), auditory thalamus and secondary auditory cortex using activity-dependent induction of immediate early gene zif268. Consistent with the reports that amygdala plays a critical role in fear memory, the zif268 activity in the LA was significantly increased after both recent and remote memory retrieval. However, the density of zif268 (+) neurons in both auditory thalamus and cortex, particularly in layers IV and VI, was increased only after remote but not recent fear memory retrieval. This finding supports that the LA is a key brain site for permanent fear memory storage and suggests that auditory upstream might play a role for remote memory storage or retrieval of auditory conditioned fear. In spite of general acceptance of the necessity of LA for fear memory, it still remains elusive whether LA is sufficient to store the fear association. The results showing the participation of auditory upstream during fear memory formation and retrieval argues that the concurrent synaptic change in those areas is also necessary. To isolate the essential neural circuit components representing fear memory association, in the second experiment I tested whether direct activation of presynaptic sensory inputs in LA, without the participation of upstream activity, is sufficient to form fear memory in mice. Photostimulation of axonal projections from the auditory thalamus and cortex was paired with aversive footshock. Twenty-four hours later the same photostimulation induced robust conditioned freezing, directly proving that synapses between sensory input areas and the LA actually are sufficient to serve as a conditioned stimulus for long-term fear memory formation. However, in the further study, I found that the fear memory trace lacking upstream circuits of LA could not support remote memory retention. The optogenetically induced fear memory lasts a few days and gradually decayed through time passage by 20 days. Surprisingly, this remote memory impairment could be fully rescued if the memory is reactivated in a few days. The reactivation was also effective even when the stimulus was given in anesthetized state where subjects cannot consciously and behaviorally experience the event. In contrast, re-exposure to the conditioned context failed to rescue the remote memory. This result suggests that a newly formed fear memory needs a reactivation process of the memory circuit to become persistent. In the third experiment direct activation of memory neuron in LA is tested. The previous study showed that LA neurons expressing elevated levels of cAMP response-element binding protein are preferentially recruited into fear memory traces and are necessary for the expression of those memories. However, it is unknown whether artificially activating just these selected neurons in the absence of behavioral cues is sufficient to recall that fear memory. Using an ectopic rat vanilloid receptor TRPV1 and capsaicin system, I found that activating this specific ensemble of neurons was sufficient to recall established fear memory. Furthermore, this neuronal activation induced a reconsolidation-like reorganization process, or strengthening of the fear memory. Thus the findings establish a direct link between the activation of specific ensemble of neurons in the lateral amygdala and the recall of fear memory and its subsequent modifications. All together, this study provides direct evidence that the LA is sufficient to encode and store associations between sensory cue and aversive stimuli and suggests new idea that learning-related synaptic change in auditory upstream plays a role for the memory persistence by systemically reactivating the memory neurons in LA. This reactivation of the memory circuit may underlie the mechanism in which a memory is properly recalled and modified according to the constantly changing environments.