Association learning drives synaptic plasticity at feedforward synapses in somatosensory cortex.

Area of Science:

• Neuroscience
• Synaptic Plasticity
• Sensory Cortex Function

Background:

• Neocortical synapses undergo plasticity during learning, but input and target specificity remain unclear.
• Feedforward synapses in sensory cortex have critical periods and are thought to resist later plasticity.
• The impact of learning, especially causal association tasks, on these feedforward synapses is largely uninvestigated.

Purpose of the Study:

• To investigate whether feedforward synapses in the barrel cortex can be altered by training in a whisker-dependent association task.
• To determine the role of reward-prediction accuracy in driving synaptic changes in sensory cortex feedforward pathways.

Main Methods:

• Pathway-specific optogenetic stimulation in freely-moving mice.
• Analysis of quantal excitatory postsynaptic currents (EPSCs) in layer 2/3 (L2/3) pyramidal neurons.
• Comparison between true training, pseudotraining (decoupled stimuli/rewards), and environmental enrichment.

Main Results:

• Training induced rapid, transient potentiation of layer 4 (L4) to L2/3 inputs, without altering thalamocortical inputs to L4.
• Pseudotraining led to depression of L4-L2/3 quantal EPSCs.
• Environmental enrichment did not affect quantal EPSC amplitude, suggesting reward prediction accuracy is the key factor.

Conclusions:

• Feedforward plasticity in primary sensory cortex is driven by reward-prediction accuracy during goal-directed tasks.
• Sensory circuits exhibit distinct synaptic changes (potentiation/depression) based on stimulus-reward contingency.
• These findings highlight the sensitivity of feedforward sensory pathways to the predictive value of sensory information.