The role of subspaces and cell types in cognitive and motor processes
OA Version
Citation
Abstract
One of the most fundamental functions of the brain is to produce movements. In mammals, the motor cortex (MCtx) supports flexible behavior by coordinating a diverse set of processes. Neurons in the MCtx produce motor commands, as well as activity related to movement preparation, motor learning, and feedback control, amongst other cognitive processes. These processes are often multiplexed in the same local circuit and cells. How can a single brain region support distinct processes while ensuring they do not interfere with each other? A prominent hypothesis proposes that neural activity separates cognitive and motor processes at the population level by producing distinct activity patterns — some that evolve in in ‘output-null’ subspace and cancel at the level of the muscles, and some that evolve in an ‘output-potent’ subspace that drive changes in muscle activity. This mechanism provides a cell-type-anonymous explanation for how only specific signals culminate in motor output. However, the MCtx contains a diversity of cell types that form unique local and long-range circuits, providing an anatomical substrate for the separation of distinct functions. It remains unclear if activitypatterns captured by these subspaces reflect the dynamics of specific cell types.
This dissertation addresses this question by relating the activity of medulla- projecting pyramidal tract neurons (PTN) in the mouse MCtx to the neural activity captured by null and potent subspaces. Using a behavioral paradigm that involves multiple cognitive processes, we show that dynamics commonly taken to support cognitive processes are strongly contaminated by movements. When cognitive and motor components are isolated using a novel approach for identifying null and potent subspaces, we find that they exhibit distinct dynamical trajectories. Using optogenetic tagging to perform cell-type-specific electrophysiology, we find that both PTNs and other neurons in the local circuit contain subpopulations that contribute selectively to either subspace, but generally not both. These findings demonstrate that distinct neural activity patterns generated by unique subpopulations—and not the activity of a specific cell type alone—enable the motor cortex to multiplex distinct cognitive and motor processes. This work provides a general framework for separating cognitive and motor processes and highlights the importance of population-level analyses in understanding cortical function.
Description
2025
License
Attribution 4.0 International