Motoneurons do what motoneurons have to do.

Abstract

How are coordinated movements possible? In terrestrial vertebrates, spinal motoneurones control movements of the limbs and the trunk. Motor units by the thousands can in theory provide almost infinite degrees of freedom, which may be good for flexibility but makes for computational complexity. It would not be surprising if billions of neurones participated in premotor processing, but if so wouldn't motor chaos reign in such a system? We presume it does not, but why not? A key factor seems to be the intrinsic response properties of the spinal motoneurones themselves. These properties should allow motoneurones to fire when they ought to and hold fire at other times. And when they fire, intrinsic properties make them fire the way they should, in terms of frequency and duration. Can this really be a true picture? First, like it or not, all limbed terrestrial vertebrates share fundamental problems of motor coordination and are bound by similar biomechanical constraints. For some time it has been clear that these are paralleled by corresponding intrinsic response properties in spinal motoneurones in different species. Not only are current-frequency relations and adaptation patterns practically indistinguishable in frog, turtle and cat, but the underlying molecular mechanisms also seem to be the same. Early on it was recognized that these firing properties ensure that any change in firing rate also results in a change in contraction force. In this way intrinsic properties optimize the relationship between synaptic input to motoneurones and force output in muscles. Second, the same motoneurones can have different roles at different times. It turns out that the excitability of motoneurones is regulated by the state of a particular calcium channel, CaV1.3 (L-type calcium channels expressing the α1D subunit). This channel mediates the persistent inward current, Ii, underlying plateau potentials. At rest, spinal motoneurones maintain a membrane potential far from the spike threshold. Only robust synaptic depolarization elicits spiking, thus debarring random synaptic fluctuations from generating ‘motor noise’. Under these conditions CaV1.3 is in a ‘reluctant’ state and makes no contribution to the response properties. In contrast, when CaV1.3 is facilitated via metabotropic receptors activated by synaptically released serotonin, glutamate, acetylcholine or noradrenaline, activation of Ii is promoted and excitability increased. In this state, motoneurones are easily recruited by synaptic excitation and readily generate spike trains appropriate for movements. This picture has emerged over 35 years of experimentation in isolated preparations. But does it hold completely? It remains unclear what Ii itself contributes specifically to motor function and how this contribution is regulated during real motor behaviour. It has not been feasible to address these questions in animal experiments in vivo so the question naturally arises, do the results even apply to human beings in the first place? Collins and collegues have provided some tentative answers and the prospects of exciting new avenues of investigation, in an article published in this issue of The Journal of Physiology (Collins et al. 2002). Their experiments on force development during stimulation of muscle afferents in normal human subjects suggest that CaV1.3 is present in motoneurones in man and may readily contribute to normal motor activity. Motor unit activity in human beings is compatible with the presence of Ii in human spinal motoneurones (Kiehn & Eken, 1997; Gorassini et al. 1998). In this issue, Collins et al. (2002) provide convincing, albeit indirect, evidence for this. They show that stimulation of low threshold muscle afferents from a flexor or extensor muscle in the leg gradually recruits extra force from the same muscle over a time period of tens of seconds. The force recruitment is due to increased activity in the spinal motoneurones innervating the muscle. This is compatible with synaptic facilitation of Ii. The authors also show that the properties and dynamics of this extra force are consistent with all we know from animal experiments about the behaviour of Ii, including sustained activation, hysteresis and depolarization-induced facilitation (wind-up and warm-up; Fig. 1). Apart from a biochemical or immunological confirmation of the presence of CaV1.3 in human motoneurones, what is now needed is the development of protocols to monitor the state of Ii during motor behaviour. This is virgin territory and many questions spring up. Is CaV1.3 evenly present in all spinal motoneurones? Is the state of CaV1.3 modulated over tens or hundreds of milliseconds, or rather over seconds or minutes? Is regulated expression of CaV1.3 part of motor learning and motor adaptation? Are any distinct pathologies related to the channel (Bennett et al. 2001)? Of course once again, what is it that Ii does for frogs and humans alike?