Abstracts

Rationale: Memories for emotional events tend to persist, whereas neutral events are often forgotten. This prioritization of emotional events is mediated by the amygdala, which modulates memory consolidation processes in the hippocampus and related medial temporal cortices. The present study examined whether brief electrical stimulation to the basolateral amygdala (BLA) reliably enhances memory in epilepsy patients without eliciting an emotional response. Methods: We recruited 30 epilepsy patients undergoing intracranial EEG (iEEG) with depth electrode contacts placed in the BLA and sub-regions of the medial temporal lobe (MTL). During continuous iEEG, each participant was presented a series of photographs of neutral objects, some of which were followed immediately by stimulation to the amygdala (8 trains of 50-Hz pulses). In further experiments, we manipulated stimulation amplitude (0.5-3.5 mA), duration (1-3 sec), stimulation timing, and location (BLA or MTL sub-region) across subsets of patients to determine the stimulation parameters for optimal amygdala-mediated memory enhancement. Recognition memory for half the images was tested immediately after the study session and for the other half of the images the following day. Free recall memory for all of the images was also measured just before the immediate recognition test. Results: Across all patients and stimulation parameters, participants recognized neutral objects initially followed by amygdala stimulation more accurately than control objects during the one-day test. No epileptiform activity was elicited by the stimulation at any point during the experiment. Participants reported no awareness of the stimulation for any stimulation condition. At an immediate delay, patients freely recalled more items previously paired with stimulation. Notably, patients with greater cognitive impairments due to epilepsy showed the largest gains in memory from amygdala stimulation at the one-day test. This memory enhancement effect was demonstrated in patients that had clinical seizures between encoding and test, as well as in patients with hippocampal seizure onset foci ipsilateral to the stimulated amygdala. Greater stimulation amplitudes and durations produced similar memory enhancement effects to the original stimulation parameters during the one-day test. Stimulating other MTL subregions, like the hippocampus, with the same parameters as prior BLA stimulation experiments produced an impairment of memory for stimulation-paired objects. In addition, accurate recognition of stimulated objects elicited oscillatory activity between the BLA, hippocampus, and perirhinal cortex that resembled the prior stimulation pattern. Conclusions: Thus, human amygdala stimulation prioritizes event-specific declarative memories in long-term storage, the recollection of which is associated with a distinct electrophysiologic signature of prior stimulation. With further tuning, amygdala stimulation may provide a novel therapeutic intervention for memory disorders, like those impairments commonly found in epilepsy patients, through selective enhancement of specific memories. Funding: NIH Grants KL2TR000455, R01NS088748, and R01MH100318 and National Research Service Award F30 MH095491