Primary and supplementary cortex in bimanual movements: a study of cortical physiology

The study of bimanual coordination is an area of intense recent research. This contrasts with a relative lack of interest in bimanual coordination – particularly in the area of cortical physiology – over the preceding years. This increased recent interest has been fueled in part by new theoretical approaches and recording techniques that place an emphasis on the role of neuronal interactions in the function of the nervous system and partly by an increasing emphasis on natural motion in the study of motor control. The central issues addressed by this research regard the nature of cortical involvement in coordinating bimanual movements, the relative role of different cortical areas, and the nature of the interhemispheric interaction underlying coordination. This study of the neurophysiology of bimanual control focuses on the relative roles of primary motor cortex (MI) and supplementary motor cortex (SMA). MI is classically considered to play little role in bimanual coordination, while SMA has been thought to have a large role. We adapted the standard center-out reaching task used in many motor control and motor systems physiology experiments so that it was suitable for bimanual research. The subjects (rhesus monkeys) held one two-joint X-Y manipulandum with each hand. They first brought both hands to starting positions located roughly in front of each shoulder. Then the monkey was instructed to make either a unimanual or bimanual movement. Bimanual movement involved either parallel movements or anti-parallel movements of the hands, and both bimanual and unimanual movements could be in any of eight directions. During performance of this task, neuronal activity was recorded by four microelectrodes in each hemisphere in MI and SMA. As expected, both in MI and SMA neurons showed activity during unimanual movements related to both contralateral and the ipsilateral movements. Also as expected, a slightly larger percentage of MI units responded contralaterally while more SMA units responded bilaterally. However, in contrast with our expectations, we found many units in both MI and SMA whose activity during bimanual movements was different than during unimanual movements. This surprising finding of “bimanual related” activity in units in MI proved resistant to explanation by differences in the kinematics or dynamics of the movements performed by the monkeys. It also turned out that, for many neurons, activation in bimanual movements could not be explained as a linear sum of activation during the unimanual movements. The percentage of MI neurons exhibiting such “bimanual related” activity was no lower than the percentage of SMA neurons, which is a strong indication for a role for MI in the coordination of bimanual movements. To pursue the source of “bimanual related” components in single unit responses, we examined the relations between local field potentials (LFP) and single-unit activity. Like single neurons, activity in the LFP could be evoked in relation to movements of the contralateral and ipsilateral arms. However, in comparison with the single units, MI proved to be driven much more strongly by contralateral movements than SMA. Another difference between the LFP and single unit activations was that “bimanual related” activity in the single units could involve either an increase or a decrease in the spike rate, while LFP responses evoked on nearly all electrodes was greater during bimanual movements than during unimanual movements. This difference in the “bimanual related” response may reflect a greater overall cortical activation during bimanual movements, which would mean an increased synaptic drive to the neurons in both motor areas. However, it remains to be explained why this increased synaptic drive would not produce an increase in the average firing rate. Another interesting difference between SMA and MI arising from analysis of the LFP is that there is a greater correlation between the responsiveness of neurons recorded in SMA and the size of the LFP evoked on the recording electrode than there is in MI. That is, SMA neurons seems to be more closely related to the LFP than MI neurons potentially indicating that a larger percentage of the drive to these neurons is local. While it is premature to draw any far reaching conclusions from the research presented it here, there seems to be justification to say both that MI is involved in the control of bimanual coordination and that it may nevertheless be true that MI and SMA play different roles in the control of movement. In any case, this research will certainly force a re-examination of the role of MI in motor control and could prove a springboard for further research which explores the physiological basis and functional relevance of the “bimanual related” effect. Opher Donchin, Doctoral Thesis Cortical Representations of Bimanual Movements