Yoshio Takashima

Yoshio Takashima, Ph.D.
Assistant Project Scientist

Current Research

Investigate the underlying mechanisms of experience-dependent plasticity in voluntary movement encoded by the motor system.

Motor learning is a process of improvement in an execution of intended movements with training over time. Acquiring motor skills involves multi-complex motor systems integrating and processing various inputs to plan a sequence of voluntary movements, activating the musculatures involved in movement execution, and to complete a task with speed, accuracy, and consistency. Among all the motor skills rodents can possibly learn and perform, act of skilled forelimb-grasping involves implementation of both reaching and grasping movements to retrieve an object such as food. Reaching employs a sequential coordination of both proximal and distal musculatures, including the muscles of the shoulder, arm, and hand. Meanwhile, grasping employs distal musculatures coordination controlling the forearm, wrist, and digits.

Our work in the Tuszynski lab investigates how a new adults motor circuit is constructed when an adult rat learns a new motor skill. In studies to date, we and others have shown adult rodents are capable of learning the skilled forelimb-grasping motor task. Within two weeks, grasping accuracy – which is indicated by percentage of success over total attempts – increases up to 70%. Moreover, we have identified distinct modifications in adult motor circuitry as a function of learning a new skilled-motor task. These learning-induced changes include augmentation of neuronal structures (spines and dendrites) and complexity of projections, increase in intrinsic excitability and recurrent connectivity in task-relevant sub-population of layer 5b (L5b) corticospinal neurons (CSNs) in the primary motor cortex (M1).  Moreover, our lab's work has shown that thalamic projection to the M1 is capable of experience-dependent plasticity after skilled forelimb-grasp learning. There is a significant increase in the amplitude of evoked excitatory post-synaptic currents (eEPSCs) propagating from thalamus to task-related sub-population of L5b CSNs after training. Notably, the change occurs specifically in thalamocortical (TC) input onto CSNs projecting to distal musculatures that are required to execute the movement in grasping sequence. There are no concomitant changes in eEPSCs onto CSNs that project to proximal musculatures. Our results demonstrated a highly specific role in plasticity of TC projections during adult motor learning.

As suggested above, several neural systems that modulate motor control come together in M1 to execute final, skilled-motor acts. These include associational inputs from the frontal cortex, parietal cortex, contralateral motor cortex, as well as subcortical input arising from the thalamus. In addition, subcortical modulatory systems influence cortical motor circuits, including cholinergic, dopaminergic, and noradrenergic inputs. When animals acquire a new motor skill, they integrate and process these various inputs to assemble a new motor program at M1 where coded information is transmitted through output neurons including CSNs located in L5b. While these processes are not completely understood, understanding these newly formed neural ensembles in the adult brain is fundamental to appreciating the nature of learning: how does the nervous system integrate inputs across different regions of the brain to interpret and code program to generate a newly learned behavior?

We hypothesize that changes in circuit connectivity, synaptic efficacy, and intrinsic excitability of neurons within M1 contribute to the changes in neuronal responses after skilled-motor learning in behaviorally- and functionally-relevant neuronal sub-populations. Most recently, our lab showed that not only enhancement in intrinsic excitability within the behavior-relevant sub-population of CSNs, but also increase in network connectivity among those neurons. Here, my work focuses on tans-laminar input propagating from layer 2/3 onto task-relevant sub-populations of L5b CSNs in M1. In order to investigate the role of L2/3 inputs in relation to skilled-motor learning, we employed viral tools, electrophysiology, in utero electroporation, and behavioral assays. Our results reveal that input from the L2/3 neurons onto behaviorally- and functionally-relevant CSNs is involved in temporal precision inputs which allows the animal to coordinate sequence of intended movement necessary to execute the trained motor task with fast sequence of action with high precision.