Interactions between motor cortex and dorsolateral striatum in a model of naturalistic motor behavior

The National Institute of Neurological Disorders and Stroke awarded The Trustees of Columbia University in the City of New York $150,932 on March 1, 2026, under the Extramural Research Programs in the Neurosciences and Neurological Disorders program (CFDA 93.853) to develop computational and theoretical models characterizing interactions between motor cortex and dorsolateral striatum in naturalistic motor behavior.

The project uses simultaneous multi-region neural recordings from mice combined with motor-syllable-segmented behavioral data and theoretical neuroscience tools to construct a framework explaining how motor cortex and dorsolateral striatum work together to generate and sequence natural behaviors such as exploration, locomotion, climbing, and rearing. The research applies computational modeling and theoretical analysis to characterize and compare neural representations of motor actions across these brain regions, disentangle their inter-regional interactions, and construct a multi-compartment model of their coordinated function. By explicitly modeling the division of labor between these regions, the framework addresses why modularity is critical for flexible motor control and provides a tool for guiding future experiments.

Work is performed at Columbia University's Health Sciences Division in New York, New York. The period of performance extends from March 1, 2026, through September 30, 2028. This is a Project Grant, a direct assistance type under NINDS's extramural research program supporting basic neuroscience research.

Abstract

Natural behaviors, such as exploration, are composed of distinct, highly stereotyped motor syllables-like locomoting, climbing, and rearing-that are seamlessly sequenced and executed by a network of brain regions. Understanding natural behaviors first requires an understanding of how these motor syllables are represented, transformed, sequenced, and executed through cross-region interactions. However, due to technological limitations, previous studies of motor control have largely examined individual brain regions in isolation and relied on highly constrained behavioral paradigms, leaving open the question of how regions like the motor cortex (M1) and dorsolateral striatum (DLS), two key nodes of the cortico-striatal network, interact to generate spontaneous, natural behaviors. By combining newly available simultaneous multi-region recordings and motor-syllable- segmented behavioral data from mice with theoretical neuroscience tools, this project will develop a theory for how M1 and DLS work together to orchestrate movement sequences. In close collaboration with experimentalists, i will use computational modeling and theoretical analysis to: (1) characterize and compare neural representations of motor actions across M1 and DLS, (2) disentangle their inter-regional interactions, and (3) construct a multi-compartment model explaining how these regions work together to generate and sequence natural behaviors. By explicitly modeling the division of labor between M1 and DLS, this framework will allow me to elucidate why modularity is critical for flexible motor control. Additionally, by aligning the model with the brain's structure, IT will provide a powerful tool for guiding future experiments. This project will advance our understanding of the neural computations underlying natural movement, bridging the GAP between theoretical models and biological reality. By capturing how brain areas interact when planning and execution unfold simultaneously in a neurally constrained model, this work will generate experimentally testable predictions. In doing so, IT will not only refine our models of motor control but also provide insights that could inform new therapeutic strategies for movement disorders such as parkinson's and huntington's disease. Ultimately, this work will provide a powerful framework for understanding the distributed computations that underlie flexible, spontaneous behavior.