Neural Commands for Fast Movements in the Primate Motor System

Fast movements are important for survival, whether it be catching prey or avoiding being caught, or a pedestrian jumping out of the way of an approaching car. Fast movements are much prized in human sports. Muscles are made up of muscle fibres, which are bundled functionally into motor units. Each motor unit is activated by a single nerve fibre, which originates from a motoneuron cell in the spinal cord. There are typically several hundred motor units in a muscle. Recent work suggests that these spinal motoneurons connected to a given muscle must be activated very closely synchronised in time to maximise movement speed. Various different brain and spinal cord systems provide inputs to motoneurons which drive their activity. Surprisingly, there are few measurements of how these inputs are activated during fast movements, but the available data suggests a sluggish increase in activation, in marked contrast to the sharply coordinated and synchronous firing of motoneurons. A major unknown is how such tight motoneuron synchrony is achieved.

In this proposal, we will test the idea that motoneurons are transiently inhibited (I), before being excited (E) to generate fast movement. Our preliminary evidence, from indirect measures in humans and computer modelling, suggests that this 'I-E' activation scheme could increase movement speed by providing better motoneuron synchronisation. We will test this idea using recordings from healthy human subjects and awake behaving monkeys. This will provide theoretical insights into neural circuit function and the role of inhibition, and practically will help us to understand limits to speed performance in health and disease.

In humans, we will use fine electrodes placed within a muscle, and surface grid electrodes on the skin overlying a muscle, to record muscle activity. We will use mathematical methods to separate the multi-channel recordings into the activity patterns of single motor units. We will analyse these to look for evidence that initial inhibition speeds up movement. We will also ask subjects to undertake a 4-week period of training to increase speed, and measure how this changes the timing of motoneuron activation.

Monkeys will be trained to perform fast movements in response to auditory and visual instruction on a computer screen. We will record motor unit activity using surface grid electrodes, just as in humans. In addition, we will insert neural probes with 1024 closely-spaced recording sites into the motor cortex, reticular formation and spinal cord - three important centres which generate the input to activate motoneurons. We will extend some of the analysis methods developed for recordings from muscle to these neural recordings. This will allow us to resolve activity from many neurons simultaneously. Using mathematical methods which look at how the activity of one cell influences the activity of another, we will identify connections between cells locally, and to motoneurons. This approach also permits us to distinguish inhibitory from excitatory cells. We will use these recordings to search for evidence of 'I-E' drive, both to motoneurons and from one part of the central nervous system to another (e.g. from the cortex to the brainstem). Finally, we will exploit this rich dataset to quantify the relative importance of cortical, brainstem and spinal sources of motoneuron input during fast movement.

This project will provide fundamental knowledge underpinning our understanding of motor control at the limit of human/animal performance, using technologies which have only recently become available. Success may help us to improve individual performance, e.g. in sport, or in patients who find that their movements are slowed after damage to the motor system or with ageing. Increased understanding of the properties of 'I-E' drive, in a system ideally suited to dissect this, may reveal novel principles of neural communication applicable more generally and across species.

Feedforward connections form a key part of neural circuits, allowing onward propagation of activity. To ensure tightly synchronous activation, previous modelling and experimental work suggests that excitatory input from a distant area should be rapidly followed by inhibition ('E-I' activation). This shortens the time window for activation, and enhances synchrony across the population. However, this scheme is often not suited to motor circuits. For example, during a fast movement, spinal motoneurons must make a strongly synchronous initial discharge followed by sustained firing. Delayed inhibition would prevent sustained firing.

This project will examine whether an alternative input pattern is used. We hypothesise that the target area is first inhibited, and then excited; this produces first spike synchrony, whilst permitting subsequent spikes to occur at high rate. This hypothesis will be tested using high-density multi-channel recordings of muscle activity in humans, and muscle and neural activity in awake behaving monkeys. We will investigate how improvements in inhibition correlate with faster movement, both across subjects and also as a result of training within a subject. We will search for a pattern of 'I-E' activation in inputs to motoneurons from the corticospinal tract, reticulospinal tract and spinal cord interneurons, in local circuits within the cortex, brainstem and spinal cord, and in descending connections from cortex to brainstem and spinal interneurons. We will also quantify the relative importance of these different sources of motoneuronal drive during fast movements and tonic contractions. Our results will be of importance to motor control, improving understanding of the motor command and how it is developed and shaped by neural circuits. In addition, the principles discovered from motor pathways with well-known connectivity may have widespread application to local and long-range transmission of information within the nervous system.