Abstracts

Rationale: An episodic memory happens only once. Thus, the encoding of an episodic memory must be successful after a single experience. The medial temporal lobe (MTL) is necessary for learning as well as retrieval of such declarative memories. For example, the hippocampus is known to be crucially involved in distinguishing between novel and familiar stimuli. Since a stimuli becomes familiar after a single presentation, such a mechanism must necessarly adapt very quickly. During retrieval of familiar stimuli, we can sometimes confidently report that we have seen something (familiarity) but can not recollect where or when it was seen. Thus, the two components of episodic memory, familiarity and recollection, can be behaviorally dissociated. It is not clear, however, whether these two components of memory are represented separately by distinct brain structures or different populations of single neurons in a single anatomical structure.Methods: Intracranial depth electrode recordings from 10 patients were analyzed. Eight small high-impedance microwires were embedded in each depth electrode to allow the recording of extracellular spikes. Spike sorting techniques were used to associate each spike with a single unit. Recordings were made while patients were engaged in a learning and retrieval task. The learning task consisted of a sequence of natural pictures presented at different locations on the screen. Patients were asked to remember which pictures they had seen where. After a delay of 30min or 24h, patients were shown the same and new pictures and asked which pictures they thought were old and new.Results: Two classes of neurons that distinguish novel from familiar stimuli were found: one that increases firing selectively to novel stimuli ('novelty detectors') and one that increases firing selectively to familiar stimuli ('familiarity detectors'). We identified these neurons bilateraly both in the hippocampus and amygdala. The average firing rate difference of all such cells is 56% between new and old stimuli. The same difference is significantly smaller (73% v. 33%) when comparing units from the 'healthy' side with units from the eventually resected side. Also, the firing response is graded: the response is strongest for items which are both recognized and recollected, intermediate for items only recongized and smallest (but significant) for items which were forgotten. Conclusions: Single-neurons in the human MTL accurately signal the novelty and familiarity of stimuli. They are capable of changing their firing rate to a particular stimulus after a single-trial. This is one of the fundamental criteria for the neural basis of episodic memory. Further, the same neurons represent both the recollective and recognition aspects of a memory. Responses are clearly existent even in the absence of recollection. This data is compatible with the idea of a continuous signal of memory strength: the stronger the neuronal response, the better the memory. Sources of funding: Gimble Discovery Fund, Howard Hughes Medical Institute