List of authors and contributors include: MA Sahakian Robbins, M Corrales, AJ Correia da Silva, M Clayton, S Mohinta, A Cardona, M Zlatic, MRC Laboratory of Molecular Biology, Department of Physiology, Development, and Neuroscience, Department of Zoology, University of Cambridge, UK.
Abstract:
A SYNAPTIC LEVEL RECONSTRUCTION OF A MEMORY ENGRAM USING ELECTRON MICROSCOPY. MA Sahakian Robbins, M Corrales, AJ Correia da Silva, M Clayton, S Mohinta, A Cardona, M Zlatic, MRC Laboratory of Molecular Biology, Department of Physiology, Development, and Neuroscience, Department of Zoology, University of Cambridge, UK. INTRODUCTION: The memory engram was first proposed 100 years ago in 1923 by R. Semon. Since then, the conventional view has been that short-term memory comes from changes to molecules within synapses following posttranslational modifications, such as changes to vesicle positioning, endo- or exo-cytosis of receptor subtypes and their clustering with adhesion proteins. whereas long-term memory is hypothesized to be stored structurally by loss and formation of synapses of engram cells. However, nobody has yet been able to comprehensively characterized these molecular and structural changes occurring within a neuron – and between neurons in a network, in a learning animal to prove that they are causal to encoding a specific memory. Here, we propose to use a tractable model system – the Drosophila larva – to comprehensively map the memory engram. METHODS: Using the advanced cell specific genetic tools available in Drosophila, we form the memory engram for Pavlovian conditioning by using optogenetics to activate or image single cells of interest. At the ‘highest level’ structural changes of this memory network are mapped using enhanced focused ion beam Scanning electron microscopy (FIB-SEM). As the connectome of the entire larval brain has recently been mapped, it can be used as a reference to readily identify a specific engram that has been formed. And the compact larval brain allows rapid electron microscopy reconstruction of all of the synapses in the network from multiple animals. This will allow us to show the input specific, Hebbian nature of the memory, as all synapses – active and inactive during learning – are mapped. These engram cells are then be isolated for further transcriptomic and proteomic studies. RESULTS AND DISCUSSION: We compare differences between synapses of trained and untrained brains 2 h post-learning for two different styles of learning. We find the differential expression of genes is already significantly different at 2 h post-learning and include human disease-related genes, and therefore may offer insight into future therapeutic avenues to prevent memory loss in dementia-related disorders. RESEARCH SUPPORT: UKRI Medical Research Council and Wellcome Trust.
The talk was a part of the Neuroscience Week in Yerevan, Armenia.
Event Website
Posted on 26/05/2026
