Cortical and Sub-cortical Contributions to Bimanual Coordination

Many everyday actions require coordinated movement of the two hands. For example, when we unscrew the top of a bottle, one hand exerts torques on the cap whilst the other holds the bottle steady. The holding hand must precisely compensate for the torques imposed by the other hand, otherwise the bottle will slip. In this project, we will investigate the neural systems underlying bimanual coordination. Macaque monkeys will be trained on two tasks, involving squeezing levers between finger and thumb with both hands. In one task, forces exerted by one hand must be compensated by the other - similar to the bottle opening example above. In the other task, the relative timing of the two hands must be precisely controlled. We expect that in the first task, neural systems controlling each hand will exchange information about movement parameters (level of force); in the second, they will exchange timing information. We will then make recordings from multiple regions of the nervous system which could be involved in bimanual coordination. In the cerebral cortex, we will focus on two motor regions. Rather than just recording any cells, we will search specifically for neurons which send projections to the other side of the brain - either to the cortex, or to the basal ganglia. We will identify these cells using a stimulation paradigm, and then observe their firing during the trained bimanual movements. This is a powerful approach, as it allows us to place the natural activity of the neurons during tasks of interest in the context of their known anatomical connections with the opposite side. Secondly, we will record from the cerebellum and reticular formation. Both regions are important for bimanual control. The reticular formation on one side of the brain sends projections to both sides of the spinal cord. We have recently shown that the cerebellum can also influence movements bilaterally. Recording from these regions will indicate their relative importance for the two types of bimanual coordination (parametric vs timing) represented by the trained tasks. Finally, we will record from the region of the spinal cord believed to contain 'commisural interneurones'. These cells send axonal projections to the other side of the cord, and provide a spinal pathway by which information could be exchanged for the coordination of bimanual movement. The spinal recordings in awake behaving monkeys will be supplemented by recordings in terminally anaesthetised monkeys. In these experiments, we will record from the spinal cells which control muscles (motoneurons), and measure the inputs from the other side of the cord via presumed commissural interneuron pathways. This will provide detailed information about the importance of this system in the control of the primate hand: until now, commissural interneurons have been investigated mainly in the part of the cat spinal cord which controls the hind limb. This project will provide important new information on how the brain coordinates movements bimanually. Most previous studies in this field have used behavioural observations in humans. Previous work in trained monkeys has investigated unidentified neurons in the cerebral cortex. This project will record neural activity directly, during representative bimanual tasks. The subset of cortical neurons which project across the midline will be specifically targeted. It will also record from important sub-cortical and spinal centres. These are often ignored in discussions of bimanual control, mainly because of their inaccessibility to non-invasive methods in man. Yet the motor system is a distributed system, and all movements result from the coordinated action of multiple interconnected structures. In bimanual control, as in other areas of motor control, we will not understand how the brain achieves the exquisite performance which we all take for granted unless we investigate all the component systems.

In this project, we hypothesise that there are two distinct types of bimanual coordination. In the first, detailed parameters of movement must be transmitted from one hand from the other. In the second, information about movement timing must be transferred. We further hypothesise that these distinct types of bimanual coordination are mediated by sub-cortical and cortical systems respectively. Experiments to test the hypotheses will be carried out in macaque monkeys. Animals will be trained to perform two tasks which require bimanual coordination of force or timing. They will then be surgically implanted to record multiple single unit activity. Initial recordings will target motor cortical areas. Neurons which project axons to the contralateral hemisphere via the corpus callosum, including those with connections to the contralateral striatum, will be antidromically identified by stimulation through chronically implanted electrodes. After recording from the cortex, microelectrode penetrations will target the deep cerebellar nuclei (interpositus and dentate) and the reticular formation. Finally, a chamber will be implanted over the cervical spinal cord, and the activity of interneurons in the intermediate zone will be recorded. By measuring the activity of neurons during the trained bimanual tasks, we will determine the extent to which cortical, sub-cortical and spinal systems contribute to different types of bimanual coordination. In a separate series of experiments in terminally anaesthetised monkeys, we will examine commissural interneurons in the primate cervical spinal cord. We will characterise the connectivity of this system using stimulation of one side of the cord, and intracellular recording from contralateral antidromically identified motoneurons. By conditioning spinal stimulation with activation of descending and peripheral inputs, we will determine both the inputs and outputs of the commissural system.