BIOLOGICAL ALGORITHMS FOR LEARNING IN THE MAMMALIAN BRAIN

Abstract

The neocortical column is an exquisitely organized and highly-conserved structure composed of >30 different cell types connected in a specified and hierarchical manner. Experience-dependent plasticity is broadly observed across neocortical areas and is facilitated by the anatomical and synaptic properties of neocortical neurons, but how these diverse cell-types and patterns of connectivity generate long-lasting changes in neural response properties during learning remains unknown. Here we will use a novel, high-throughput animal training system to investigate the progressive reorganization of the neocortical circuit during reward-association learning. We will focus on alterations in inhibition from somatostatin (SST) neurons, as SST neural activity is regulated by reinforcement cues, provides strong synaptic inhibition to diverse neurons across the cortical column, and dynamically controls excitatory synaptic transmission through presynaptic GABAb receptors under basal conditions. Our preliminary data indicate that GABAergic SST synapses onto Pyr neurons undergo a profound reduction in input strength in the early phase of sensory association training. Our experiments will test the hypothesis that this disinhibition enables the generation of prolonged activity within the cortical circuit that may ultimately be necessary for stimulus-reward coupling and excitatory synaptic plasticity during learning. Results generated from the proposed experiments will provide insight into how the biological neural networks are organized and regulated to enable learning.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 12, 2021
Source ID
FA95502010134

Entities

People

  • Alison Barth

Organizations

  • Air Force Office of Scientific Research
  • Carnegie Mellon University
  • United States Air Force

Tags

Fields of Study

  • Biology

Readers

  • Neuroscience

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks