Related Experiment Video
Updated: Aug 28, 2025
06:18
Author Spotlight: Deciphering Neural Circuit Formation from Two-Photon Microscopy and Single Neuron Imaging
Published on: November 21, 2023
895
Learning enhances encoding of time and temporal surprise in mouse primary sensory cortex
Rebecca J Rabinovich1,2,3, Daniel D Kato1,2,3, Randy M Bruno4,5,6,7
1Department of Neuroscience, Columbia University, New York, NY, 10027, USA.
Nature Communications
|September 20, 2022
Summary
Reinforcement learning significantly reshapes sensory cortex representations. Previously inactive neurons become responsive, enhancing touch sensitivity and behavioral choice encoding during learning.
Area of Science:
- Neuroscience
- Sensory Processing
- Learning and Memory
Background:
- The primary sensory cortex was traditionally viewed as solely responsible for initial sensory information processing.
- Cortical activity, particularly in superficial layers, is sparse and influenced by various factors beyond direct sensory input, including task context and behavioral state.
Purpose of the Study:
- To investigate how reinforcement learning impacts neuronal representations in the superficial layers of the primary somatosensory cortex.
- To determine if learning alters neuronal responsiveness and encoding of sensory and non-sensory information.
Main Methods:
- Longitudinal imaging of neurons in superficial layers of the mouse primary somatosensory cortex.
- Training mice on an object detection task involving tactile stimuli and varying trial timings.
- Analyzing neuronal population activity and encoding of behavioral choices and temporal information.
Main Results:
- Reinforcement learning dramatically altered neuronal representations, recruiting previously unresponsive neurons.
- The neuronal population sensitive to touch and behavioral choice expanded significantly with learning.
- Cortical responses diminished with repeated stimulus presentation outside the task context.
- Training enhanced the population encoding of time passage, with timing deviations eliciting strong responses.
Conclusions:
- The superficial layers of the sensory cortex display substantial learning-dependent plasticity.
- These cortical layers are significantly modulated by non-sensory, behaviorally relevant features like timing and surprise.
- Sensory cortex function is more dynamic and context-dependent than previously understood.
