Still searching for the engram: Should we?

Fourth Sym­posium on Open ques­tions in Neuroscience

Center for Neurophysics, Physiology et Pathology

December, 10, 2014

In the Traité de l’Homme, Descartes described the brain as a complex mesh of tiny tubes along which different flowing patterns of ‘animal spirits’ could be physically realized, each corresponding to specific sensory information or motor commands. As the spirits flowed through, they caused persistent structural modifications of the tubes, that would in turn change the pattern in which the spirits would later flow, and – Descartes concluded – “c’est en quoi consiste la mémoire” [this is what memory consists of]. Clearly wrong in the physiological details, Descartes’ account nevertheless ushered in the basic concepts that inform current research about memory mechanisms. As Lashley famously wrote: “Substitute nerve impulse for animal spirits, synapse for pore [tube] and the result is the doctrine of learning as change in resistance of synapses”. The modern notion, epitomized in Hebb’s dual-trace theory, is that memories are stored by activity-driven changes in the pattern and strengths of the connections between a sub-set of neurons in the brain (the engram), and their persistence over time results from the consolidation of those structural changes.

Recent experimental results have provided mixed support for this synaptic trace theory of memory. In our opinion, they raise in fact more questions than provide answers. Chronic imaging experiments have demonstrated that the rate of formation of dendritic spines transiently undergoes a significant increase during learning episodes in the adult animal. Moreover, newly formed spines have a higher survival probability than spines present prior to learning. These observations fit well with the idea that new memories are encoded through synaptic re-organization of neuronal networks. Rather puzzling is, on the other hand, the observation that spines are extremely plastic also in the absence of explicit learning. In addition, in-vitro studies have reported substantial network re-configuration, even when the neural activity is blocked. The consideration of the combined effects of spontaneous re-wiring (spine turnover) and changes in efficacies (fluctuations in spine size), and of their time scales, leads to a somehow surprising conclusion: the learned synaptic trace is extremely short-lived (~ days) yet the behavioral expression of the memory can be stable over years.

If memories are encoded in the pattern and strengths of the connections between neurons, how can they be stable in face of the reported substantial volatility in synaptic structures? Could non-synaptic plasticity mechanisms play a far more important role in the formation and consolidation of memories than currently thought? Are current computational models of memory formation and retention compatible with the above described experimental phenomenology or, instead, are they in need of substantial amendments? This workshop will bring together experimentalists and theoreticians to address these questions, while critically re-examining successes and problems of the synaptic trace theory of memory.



Claudia Clopath (Imperial College London), Dominique Debanne (U Aix-Marseille), Aline Desmedt (U Bordeaux), Eric Hosy (IINS Bordeaux), Yonatan Loewenstein (Hebrew University), Jacques Micheau (U Bordeaux), Simon Rumpel (Johannes Gutenberg University in Mainz)


Conference room of the Center for Neurophysics, Physiology & Pathology
Paris Descartes University, 45 Rue des Saints Pères, Paris
(3rd floor, room H335)



09:00 – 09:30 Coffee & Croissants
09:30 – 10:30 Jacques Micheau (Bordeaux)
              Engram is coming back to the front of the scene: New data but old vista?
10:30 – 11:15 Aline Desmedt (Bordeaux)
              Switching from the search of the engram to the analysis of emergent memory – or – from a “neuronal syntax” to a
“neuronal semantic” 11:15 – 11:45 Coffee break 11:45 – 12:30 Simon Rumpel (Mainz) Dynamics of neuronal circuits in the mouse auditory cortex 12:30 – 13:15 Yonatan Loewenstein (Jerusalem) TBA 13:15 – 14:30 Lunch 14:30 – 15:15 Eric Hosy (Bordeaux) The use of super-resolution imaging techniques to revisit synaptic transmission 15:15 – 16:00 Dominique Debanne (Marseille) Long-term plasticity of intrinsic neuronal excitability: Learning rules and mechanisms 16:00 – 16:30 Coffee break 16:30 – 17:15 Claudia Clopath (London) Modeling synaptic plasticity and its implications in networks 17:15 – 17:45 Closing discussion