The cortex, interneurones and motoneurones in the control of movement

The Journal of Physiology Symposium The cortex, interneurones and motoneurones in the control of movement was held in conjunction with the 7th IBRO World Congress of Neuroscience Satellite Motor Control at the Top End. Tropical Darwin, known to Australians as ‘The Top End’, was the obvious venue for a winter meeting and it proved a great success with outstanding seminars covering many facets of top-down motor control. A selection of reports in this issue of The Journal of Physiology outlines current theories and advances in this exciting area that attempts to unite theoretical models with experimental observations.

Peter Strick kicked off, showing us his map of cortical neurones making direct connections with motoneurones innervating the hand as identified by retrograde transport of rabies virus. It appears that our image of Homunculus, with his well-defined albeit surreal features, has been subjected to an expressionist deconstruction. The hand is represented over an extensive area of the primary motor cortex (M1) with output neurones as far away as the area we had expected to see his shoulder and with a sizable representation in area 3a of the primary somatosensory cortex. Never mind. Peter proposes that this intermingling of neuronal populations could be the substrate for generating synergistic patterns of muscle activity and that we could consider the motor cortex as a central pattern generator.

Stephen Scott presented his historical view of models of primary motor cortex function: the coming and going of servo-control and the re-emergence of a focus on sensorimotor transforms (Scott, 2008). How is movement encoded in the cortex – is it target location, limb or joint motion, joint torque or patterns of muscle activation? The answer proposed is that there isn't one – there are many and essentially the brain can change them on a whim. Different cortical neurones behave differently in different tasks. Time for a new concept? He argues that a theory of optimal feedback control gives sensory feedback the flexibility essential for sophisticated volitional motor control that can adjust motor commands based on distance, movement and load, and rapidly learn the sensorimotor associations for different loads.

The direct corticospinal projections of the ventral premotor cortex (PMv) could provide the motor output for its known role in the visual guidance of grasping. Alternatively, the motor output could come via cortico-cortical connections with M1 delivering the corticospinal output. Roger Lemon showed us how hand movements produced by intracortical microstimulation of a PMv area were almost abolished by microinjections of a GABA agonist into M1. The explanation is that, at least in the acute phase, motor output from the premotor cortex relies on corticocortical activation of M1 rather than its direct corticospinal projections. How else might we explore these corticocortical connections? Transcranial magnetic stimulation (TMS) is a favourite toy to play with the human motor cortex because output to muscles is so easily measured. What about the rest of the cortex where output to other areas of the brain is not easily seen? John Rothwell told us about new ways of combining TMS and fMRI to see these connections. Most impressive was a simultaneous application, achieved in a 3T field, showing TMS responses in cortical and subcortical areas changing patterns of connectivity according to task.

What of descending influences onto spinal motoneurones? The report of Chris “CJ” Heckman considers how monoaminergic projections from the brainstem affect motoneurone excitability by generating dendritic persistent inward currents (Heckman et al. 2008). These PICs greatly amplify other synaptic input during motor activity so that even small force output would be difficult without PIC assistance. The problem is how to get specificity from this diffuse projection across many motoneuronal pools including those of antagonist muscles. Focused spinal circuits (e.g. Ia reciprocal inhibition) might achieve this because of the sensitivity of PICs to inhibition. CJ proposes that descending commands to motoneurones could be coupled to both brainstem monoaminergic nuclei and inhibitory spinal interneurones to produce the overall excitability for the larger behaviour as well as the necessary specificity for movement.

These monoaminergic influences on dendritic currents and plateau potentials were explored by Jean-Francois Perrier (Perrier & Cotel, 2008), who wondered why inhibitory influences rather than excitatory PIC effects were sometimes reported. Extracellular addition of serotonin to a slice preparation would be either excitatory and lead to plateau potentials, or inhibitory and prevent plateau potentials. In the presence of an electric field that depolarized distal dendrites, it was always excitatory and produced plateaus. This dichotomous effect was investigated by focal application of serotonin agonists either close to distal dendrites or close to the soma. Distally it promoted plateau potentials whereas at the soma it induced hyperpolarization. There is a view that the facilitation produced by plateau potentials is physiologically important for generating the tonus of antigravity muscles. If this is generated by diffuse descending monoaminergic drive onto the dendrites of motoneurones, we have to ask about the role of monoaminergic perisomatic inhibition.

Trevor Drew considered the interaction between descending corticospinal signals and the spinal networks that shape muscle activity during walking (Drew et al. 2008). Looking at how cats avoid obstacles when walking, he identifies by cluster analysis a sparse array of multiarticular muscle activation synergies that are recruited sequentially and transiently through the phases of the step cycle. Smooth limb trajectories observed when avoiding different obstacles result from coordinated changes in these synergies. Recordings from the motor cortex show that pyramidal tract neurones show similar activity during gait modifications with different groups active sequentially and transiently throughout the swing phase of the modified step. The concept developed is that the motor cortex regulates locomotion by controlling the timing and level of sparse, multiarticular muscle synergies.

Focusing on the cord itself, Véronique Marchand-Pauvert summarized her amazing human electrophysiological investigations into how spinal circuits integrate descending commands and sensory afferent signals to contribute to the postural control of the wrist during hand movements (Marchand-Pauvert & Iglesias, 2008). Be warned! This tour de force to untangle the complex network of different sensory afferents, excitatory and inhibitory interneurones both at the segment and rostral in the cervical propriospinal system, driving agonist and antagonist pairs, is not for the faint-hearted or impatient experimenter. The significance of these connections for dystonic conditions is investigated in patients with writers' cramp. Altered proprioceptive reflexes mediated through spinal interneurones and associated with abnormal wrist posture were observed in this dystonia but it is yet to be determined whether this is a causal or adaptive mechanism.

Finally, we heard about the drive to motoneurones from brain stem respiratory centres. Jack Feldman gave a polished account of the evidence that two mammalian brainstem oscillators contribute to the respiratory rhythm. The inspiratory preBötzinger Complex and the expiratory parafacial respiratory group, which normally produce a coupled output, can be differentiated by anatomical lesions, opioid sensitivity and responses to lung inflation. Of clinical significance is the observation that relatively small reductions in the preBötzinger neurone population produce a respiratory pattern during sleep resembling the apnoeic pattern observed in ageing and neurodegenerative diseases. The next step was to consider how these brainstem respiratory centres activate motoneurones and integrate respiratory and non-respiratory demands. Peter Kirkwood brought us back to motoneuronal PICs and plateaus as amplifiers of descending motor command signals. Hindlimb motoneurones show respiratory central drive potentials. This sparse connection from an unknown medullary source is greatly amplified by modulated PICs so that plateau potentials are evoked easily. In contrast, the central respiratory drive potential seen in phrenic motoneurones probably has a large respiratory bulbospinal drive and is not amplified significantly by PICs so that plateau potentials are not normally seen. Motoneurones supplying the thoracic wall muscles behave similarly but are also subject to additional tonic inputs not seen in diaphragmatic motoneurones.

Jane Butler gave an elegant presentation of her single unit work on the motoneuronal output to different human inspiratory muscles. Using a neat time and frequency analysis of unit firing properties across the respiratory cycle, she showed how diaphragmatic units are recruited much earlier in the cycle and reach higher firing frequencies than intercostal units. This is despite the intercostals having large numbers of tonically firing units and therefore being at threshold, and the diaphragm having no tonic units. This tells us that this motor drive must be organized, even within a motoneurone pool, at a premotoneuronal level. Her review paper (Butler & Gandevia, 2008) argues that we need to expand the view that recruitment is organized by motoneurone threshold (i.e. the size principle). For respiratory and probably other motor tasks, the descending drive should be seen as specifying patterns of motoneurone recruitment.

At the end of a long day, Doug Stuart entertained with one of his renowned scholarly discourses. Proclaiming that he alone in the room had the chronological credentials to muse on these matters, Doug reminisced on the prewar (the Great War!) work of Thomas Graham Brown and his proposal that mutually inhibitory spinal half-centres form an independent oscillatory network that drives locomotion. Fifty years out of time and unregarded, his work was rescued from physiological oblivion by Anders Lundberg and friends who provided the compelling evidence for the interneuronal circuity of this spinal organization.

With a warm glow endowed by this thought of eventual recognition, ‘beer-o'clock’ was chimed and all retired towards a tempting but crocodile infested beach. On a long hot day in the city holding the world record for beer consumption, the last speaker might have expected to be a lone witness. That it remained a packed house to the bitter end attests to the quality of the presentations. Thank you to the outstanding speakers and the expert Chairs of Simon Gandevia, Roger Lemon and Doug Stuart. The interested reader should look at the review and research papers that are relevant to corticospinal and propriospinal control of motoneurones and muscle activation, which are published in the present volume (Butler & Gandevia, 2008; Chew et al. 2008; Drew et al. 2008; Heckman et al. 2008; Marchand-Pauvert & Iglesias, 2008; Martin et al. 2008; Perrier & Cotel, 2008; Scott, 2008).