Resources
Authors & Affiliations
Matteo Saderi, Ankit Awasthi, Sheena Josselyn, Paul Frankland
Abstract
Memories are thought to be encoded in the brain via synaptic strengthening among memory engram cells. Still, it has not been possible to observe if learning potentiates synaptic connections and simultaneously show that these changes are causally related to memory expression (i.e., they occur in engram neurons). We used a novel engram allocation-based approach to observe synaptic changes in engram neurons upon learning in-vivo. Using optogenetics, we assigned selected Medial Entorhinal Cortex (MEC) neurons to a context fear memory. This takes advantage of the fact that neurons with more excitability are preferentially allocated to engrams. We expressed excitatory and inhibitory opsins in MEC and implanted an optrode above MEC terminals in the Dentate Gyrus (DG). Increasing the excitability of MEC terminals in mice that underwent fear conditioning allocated the fear memory to opsins-expressing neurons. In a separate cohort of animals, we used in-vivo electrophysiology to record pre- and post-learning strength of MEC to DG allocated neurons’ synapses. Synapses originating from allocated MEC neurons were potentiated 24-hours after learning. Finally, to assess a causality between engram-specific synaptic potentiation and memory, we used optogenetics to induce synaptic depression 24-hours after learning in the terminals of allocated neurons. We observed that synaptic strength reversed close to pre-learning levels. Additionally, when animals were tested again in the training context, they showed decreased freezing behavior (i.e., weaker memory) compared to control animals. Using a within-subject in-vivo design, this work supports the hypothesis that engram-specific synaptic potentiation is a key mechanism for memory encoding and storage.