Ho Ling Li; Post Doc, EPFL Lausanne (working with Prof Wulfram Gerstner)
A unified neural network model for reconsolidation and extinction of fear memory
Understanding the neural processes involved in extinguishing fear memory is critical as it sheds light on potential treatments for anxiety disorders, such as phobias and post-traumatic stress disorder. In rodents, extinguishing a conditioned fear response depends on the duration of re-exposure to the relevant fear-conditioned context. A long re-exposure leads to the loss of the fear response by forming a new extinction memory. Contrastingly, a short re-exposure causes labilisation of the fear memory, followed by a protein synthesis dependent process called reconsolidation, resulting in the persistence of the fear response. Extinction and reconsolidation are mutually exclusive processes, as proven by their different requirements on biochemicals. Right now, no existing model is able to explain such phenomena.
In this talk, I will introduce our on-going work on developing a neural network model that describes the two processes. Our network includes components that represent hippocampus for input of fear-conditioned context, medial prefrontal cortex for modulation of fear learning and expression, and the amygdala for fear response. The synaptic connections between the structures are categorised into two kinds: connections formed during fear conditioning or during extinction.
Each type of connection has its plasticity distinctly regulated by a competitive molecular mechanism. Our model is able to reproduce the temporal signatures of reconsolidation and extinction, as well as the protein synthesis dependent effects on the two processes. This work represents the first attempt to explain the underlying mechanisms of time and protein synthesis dependence of reconsolidation and extinction.
Informatics Forum, Rm 1.16