Interactions between long- and short-term synaptic plasticity transform temporal neural representations into spatial.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Synaptic Plasticity

Background:

• Neural information processing relies on synaptic transmission, with efficacy influenced by recent firing history (short-term plasticity).
• The interplay between short-term and long-term synaptic plasticity and its impact on neural network learning remains largely unexplored.

Purpose of the Study:

• To investigate how long-term modifications of short-term synaptic plasticity affect neural learning capabilities.
• To determine if neurons can learn to process temporal spike sequences as spatial patterns through plastic short-term plasticity.

Main Methods:

• Development of a theoretical model incorporating long-term changes to short-term synaptic plasticity.
• Analysis of the model's ability to learn and process temporal spike sequences.
• Comparison of model predictions with electrophysiological data from mouse and human neocortex.

Main Results:

• Long-term plasticity of short-term plasticity enables neurons to learn temporal sequences, treating them as spatial patterns.
• This mechanism enhances neural circuit capacity and robustness, albeit with increased spiking activity.
• Neurons with plastic short-term plasticity can discriminate inputs based on spatiotemporal spike correlations.

Conclusions:

• Modulating short-term synaptic plasticity via long-term mechanisms offers a novel pathway for neural learning.
• This plasticity confers flexibility in neural information processing, adapting to temporal and spatial input features.
• The study provides a theoretical framework consistent with experimental data and predicts activity-dependent learning rules.