Human hand function: the limitations of brain and brawn
Human hand function: the limitations of brain and brawn at the 2011 Physiological Society meeting held in Oxford was expected to tackle some of the roadblocks within the brain right down to the muscle.
Abstract
Since the dawn of modern neurology in the late eighteenth century, understanding how the human hand is controlled has remained a foremost challenge because of the clinical reality that hand function is a major victim of stroke and recovers very poorly. For those who survive stroke, improvement of current rehabilitation strategies is the hope for improved functional outcomes (Wolf et al. 2006). Regrettably, our knowledge of the mechanisms underlying functional weakness, flexor spasticity and muscle contracture is woefully deficient. With this background, The Journal of Physiologyhosted the symposiumHuman hand function: the limitations of brain and brawnat the 2011 Physiological Society meeting held in Oxford. The five speakers were expected to tackle some of the roadblocks within the brain right down to the muscle. Each contributor illuminated one facet of the problem and suggested some current views that need to change. Simon Gandevia reminded us that human hand function has limitations at more than one level (van Duinen & Gandevia, 2011). Well-known peripheral mechanical factors include the links between tendons that couple unintended movements of adjacent digits. Some largely forgotten limitations mean that the fingers move through joint angles at which flexion and extension of the distal finger joints are impossible due to muscle–tendon biomechanics. Central limitations, probably at a cortical level, arise not only through neural ‘spill over’ to nearby muscles but also through a process that puts a brake on force generation during strong contractions. Turning to an evolutionary view, we heard that human hand structure and function should be understood not as a simple progression from the great apes, but from an earlier hominid. More than four million years ago, Ardipithecus ramidushad a surprisingly modern hand with a relatively long and robust thumb and even an opposable little finger (Lovejoy et al. 2009). Its wrist, carpals and metacarpals allowed hyperextension but not the flexion that is associated with knuckle walking in apes. This new anthropological information is a cogent reminder that anatomical structures and potentially the neural control systems of our nearest living genetic relatives, often assumed pertinent to human function, are not necessarily extant inHomo sapiens. More than 30 muscles move the human hand through the vast array of postures used for the countless tasks it performs. Hand movements and tasks are complex and often require contraction of many or most of these muscles. Each muscle has its own functionally appropriate architecture with two common principles of motor unit organisation: small motor units are recruited before large ones and large motor units fatigue more than small units (Fuglevand, 2011). Unlike the cat, motor units of human intrinsic hand muscles cannot be identified by twitch dynamics as fast and slow units. Andrew Fuglevand presented data for correlated firing between pairs of motor units during weak voluntary contractions of hand muscles. Strong correlation with a narrow temporal peak signifies activation of two motoneurons by the same corticospinal (or other) descending axon. This measure of common input strength is high between functionally linked muscles such as flexor pollicis longus and the index finger part of flexor digitorum profundus, but low between individual intrinsic muscles, presumably indicating greater capacity for independent control of the intrinsic muscles. Further work will reveal whether this association operates throughout the upper limb muscles and how it is laid down in development and later modified by use. In recent decades, the ease of transcranial magnetic stimulation of the motor cortex has seen the role of other descending motor systems, such as the reticulospinal tract, and their propensity for major influence at the spinal level downplayed or commonly ignored. In the monkey, significant recovery of upper limb and hand function occurs after pyramidotomy (Lawrence & Kuypers, 1968). What mediates this? Stuart Baker tackled the problem of descending control of the hand through pathways other than the well-known crossed corticospinal tract. Studies from Baker's laboratory have revealed monosynaptic effects of the reticulospinal system on intrinsic hand motoneurons as well as convergence of reticulospinal and corticospinal drive on spinal interneurons that probably project to motoneurons (Baker, 2011). While the monosynaptic reticulospinal effects are weaker than direct corticospinal effects, reticulospinal cells discharge during voluntary hand movements and must be considered part of the substrate for control of the hand. Previously, this system had been linked to proximal and bilateral arm movements (Davidson et al. 2007b). Now, in the monkey, the reticulospinal projection rather than the ipsilateral corticospinal projection is emerging as a system that could contribute to functional recovery of the intrinsic hand muscles and forearm finger flexors after an experimental unilateral lesion of the pyramidal tract designed to reveal potential neural mechanisms that may operate following a stroke. The concept of ‘upper motoneurons’ faithfully transmitting motor cortical activity to ‘lower motoneurons’ is one of the historical legacies in modern neurology. Marc Schieber presented compelling arguments that this concept and the view that the primary motor cortex demonstrates strict cortical somatotopy should be buried. Several lines of evidence indicate that motor cortical activity can be dissociated from the activation of motoneurons. Fetz & Finocchio (1975) first investigated this with recordings of motor cortical cells in awake monkeys and subsequent work has confirmed the lability of the linkage between the motor cortex and motoneurons, even for cells with direct corticomotoneuronal projections (Davidson et al. 2007a). Corticospinal tract stimulation in conscious human subjects shows a contraction history dependence of the motoneuronal response compatible with this view (Petersen et al. 2003). Functional imaging reveals that imagination of a hand movement, or even watching another person's movement, can activate the cortex but with little output to motoneurons (Porro et al. 1996). Schieber dissected this important disassociation, provided data from his laboratory, and highlighted its implications for controlling brain–machine interfaces through cortical activity without motor activation (Schieber, 2011). Nick Ward, the final speaker, provided a clinical perspective on the potential for hand function recovery after stroke. Combining functional magnetic resonance imaging with transcranial magnetic stimulation and measures of motor performance has provided some understanding of the cortical reorganisation that can contribute to recovery after stroke. Which areas are involved, how do they interact, and how do they drive upper-limb motoneurons? Both cross-sectional and longitudinal studies provide insight into reorganisation of secondary motor areas (dorsal and ventral premotor, supplementary, and cingulated motor cortices) and the role of the contralesional primary motor cortex (Ward, 2011) following stroke. In patients with greatest damage to fast direct corticospinal pathways, these secondary motor areas show greater activation and contribute more to performance, and this appears to involve force production in the distal musculature of the upper limb as if these areas have adopted properties of primary motor cortex. Interestingly, changes in interhemispheric cortical function with ageing in the absence of overt pathology are providing clues about the type of changes that can emerge following a stroke (e.g. Talelli et al. 2008). The symposium closed with debate of other big and unanswered questions about hand function. How do sensory systems exert their potent influence on motor learning and hand dexterity? How and where do volitional and other drives engage the motor cortical outputs that ultimately move the hand?