Portal de Periódicos da CAPES

An enduring map of the motor cortex

2008; Wiley; Volume: 93; Issue: 7 Linguagem: Inglês

10.1113/expphysiol.2007.039081

1469-445X

Roger Lemon,

Neurological disorders and treatments

The paper by Leyton & Sherrington (1917) has had a profound and lasting impact for ninety or so years since its publication. It is notable in many different ways. Leyton & Sherrington (1917) provided the first detailed proof that there was indeed localization of function within the cerebral cortex. The durability of their report probably owes most to the fact that Leyton & Sherrington (1917) were the first to establish precisely the true extent of the motor area, and to provide the first detailed 'motor map' of the primate motor cortex. In addition, they showed that surgical extirpation of the cortical tissue that, when stimulated, gave rise to movement of a particular body part, resulted in a widespread weakness and loss of use of that same body part. There was, however, substantial recovery in the weeks that followed, recovery that was not lost on lesioning either the adjacent tissue in the same hemisphere or the equivalent cortical area of the opposite hemisphere. Finally, they were able to trace the course of the degenerating corticofugal and corticospinal fibres. They observed widespread degeneration in the cervical cord after a lesion of the hand and arm cortical area and noted that after such a lesion in the chimpanzee (p. 185), 'the whole of the cross-area of ventral horn has scattered through it many degenerating fibres…', which I think is the first report of the direct cortico-motoneuronal projection, a projection whose existence was confirmed physiologically by Bernhard & Bohm (1954), and one that appears to be unique to primates (Porter & Lemon, 1993). Charles Scott Sherrington (1857–1952) of course needs no introduction (Granit, 1966). His co-author, Leyton (1869–1921), was born Albert Sidney Frankau Grünbaum in London. He changed his name by deed poll to Leyton in 1915, probably because of the difficulties for those with Germanic names living in the UK during the Great War. He studied medicine at St Thomas' Hospital and it was there that he met Charles Sherrington, whom he followed to Liverpool when Sherrington took the Holt Chair of Physiology; it was from Liverpool that the preliminary reports of these key studies first appeared (Grünbaum & Sherrington, 1901, 1903). A. S. F. Leyton was not only a skilled physiologist; he is also well known for discovering the positive agglutination reaction observed after immunization against typhoid fever (see Stewart, 1922). This classic paper was very long in gestation. After the early accounts appeared in 1901 and in 1903, further details of their experiments were given in Sherrington's Silliman Lectures (published as The Integrative Action of the Nervous System in 1906). However, the full paper was not submitted to the Quarterly Journal of Physiology until October 1916. There is a much-repeated story, probably first promulgated by John Fulton in his 1952 biography, that Sherrington delayed publication until after the death of Victor Horsley, with whom he had a very heated row. It was believed that Sherrington wanted to avoid a repeat of the disagreement in 1894 over primacy of research on pyramidal tract degeneration. Horsley, the famous neurosurgeon and co-inventor the Horsley–Clark stereotaxic method, died in July 1916 from fatal sunstroke in Mesopotamia (where he was stationed during the War). However, this account has been seriously questioned (Vilensky et al. 2003) and there must have been other reasons for the delay; these might have included the expense and difficulty of obtaining the additional animals needed for the study, and the facts that Leyton left Liverpool not long after the 1903 paper was published, to take up a Chair in Leeds, and that the the work was undoubtedly interrupted by the outbreak of the Great War. The decision to investigate the excitable cortex of anthropoid apes probably arose because the findings of motor responses evoked by cortical stimulation in animals such as dogs, rabbits and monkeys were challenged as of being of little relevance to humans (where have we heard that before?). Sherrington considered that experiments in 'higher apes', being closer to humans than any other species, might provide the answer. The initial experiment (in a chimpanzee) having provided a positive result, they decided to complete a full series of studies, and this paper summarizes a great many different experiments in a total of 22 chimpanzees, three orang-utans three gorillas and other animals, all investigated under deep anaesthesia with chloroform and ether. One of them involved an experiment in which the same induction coil was used to stimulate, simultaneously, the cortex of a cat, a monkey and a chimpanzee, and this demonstrated unequivocally that the threshold for evoking motor response was very similar in all three species. Rereading this paper, one is struck by it having all the hallmarks of a really modern piece of neuroscience: a multidisciplinary mixture of state-of-the-art electrophysiology, behavioural analysis and detailed neuroanatomy, all linked together to test a strong central hypothesis: 'the motor cortex can be regarded as a synthetic organ for compounding and re-compounding in varied ways movements of varied kinds….' The authors considered that the movements they evoked were like words in a basic vocabulary of voluntary movement, words that could be combined in myriad ways to make sentences: purposeful movements involving co-ordinated activation of many muscles in a precise spatio-temporal sequence. There are a great many key points of lasting value in this paper, some of which are discussed below. Sherrington was very clear that the nature of the motor responses was in part due to the type of stimulation used. Leyton & Sherrington (1917) employed 'faradic' stimulation (alternating current provided by an induction coil), were in favour of unipolar stimulation (using a large and remote indifferent electrode) and insisted on studying responses evoked by a brief period of stimulation at the lowest current intensities. In this way they reduced the possibility that the responses were due to physical spread of the induced current through the tissue. Stronger and longer lasting stimulation also raises the possibility that the responses are due to physiological spread of synaptic activity generated locally at the stimulation site but permeating through both short- and long-distance connections to recruit even quite remote cortical and subcortical structures. As Phillips & Porter (1977), in their elegant discussion of Leyton & Sherrington (1917) put it: 'Thus faradic stimulation is disqualified, equally with "galvanic" as tool for evoking natural function but it can be used as 'a tool for mapping the outputs that are available for selection by the intracortical activities that it cannot itself evoke.' Previous investigators, including Ferrier and Beevor & Horsley, had used relatively large currents and thus been able to evoke motor responses from the primary sensory as well as from the primary motor cortex. Leyton & Sherrington (1917) refined the approach by carefully controlling current intensity, and demonstrated that the motor area did not extend posterior to the central sulcus. They were careful to point out that even a motor zone could influence sensation, a prescient comment given that many of the descending corticospinal fibres are involved in control of afferent input (see Lemon & Griffiths, 2005; Lemon, 2008). Leyton & Sherrington (1917), describing the 'fractional quality' of the motor responses evoked, write that 'the individual movements, elicited by somewhat minutely localised stimulations, are, broadly speaking, fractional, in the sense that each, though co-ordinately executed, forms, so to say, but a unitary part of some more complex act, that would, to attain its purpose, involve combination of that unitary movement with others to make up a useful whole.' They were also keen to emphasize the complexity of some of the responses, which often comprised a '1st movement' (e.g. thumb extension) with accompanying 2nd (e.g. index finger extension), 3rd (wrist extension) or even 4th movements. They also stressed the 'functional instability of cortical motor points'; the responses were not fixed, but showed properties such as facilitation, reversal and deviation, especially when a given cortical point was retested after the adjacent area had been stimulated. Finally, they recognized that movements of the eyes were likely to be controlled differently from movements of the rest of the body, since the area giving rise to eye movements was located rostral to and distinct from the main motor cortex (see Fig. 1), in what we now call the frontal eye fields. Motor maps of the gorilla cortex A, scale drawing of the left hemisphere of one of Leyton & Sherrington's experiments on a gorilla (gorilla 1). The numbers and letters encode a wide range of different primary movements evoked by faradic stimulation. Eye movements (372–388) were generally evoked from an area further rostral from the motor cortex. Owing to lack of space, many motor effects were not plotted. B, simplified 'map' showing 'responses grouped diagrammatically', as the authors put it. Leyton & Sherrington (1917) were clearly struck by the rich variety of different movements that could be evoked from the motor area. They reported more than 400 different '1st movements' in a table that takes up 7 pages (would the current editors of Experimental Physiology allow such indulgence?). Each movement was carefully inscribed on a scale drawing of the exposed cortex (Fig. 1A). They were careful to note that the map is not fixed in nature but that responses showed instability in time (see above), and that even the boundaries between different body 'areas' (e.g. between the face and arm area) were labile. Movement of a given body part (e.g. the thumb) could be evoked from an extensive region of motor cortex, many square millimetres in area, and this representation overlapped significantly with that of other, adjacent body parts (e.g. thumb with index finger, wrist with elbow etc.). Similar findings were published 20 years later by Penfield & Boldrey (1937) for the human brain. Somehow the complexity of the experimental findings was digested into a greatly simplified 'homunculus' published after another long delay by Penfield & Rasmussen (1952). This cartoon suggested a strictly somatotopic representation within the primary motor area, M1, which bore little relation to the original experimental data. However, the homunculus provided an appealing notion of how voluntary motor control might be represented in the cortex, and has since been reproduced in every relevant textbook. It often appears alongside Sherrington's (1906) map of the gorilla cortex, which bears some resemblance to figure 10 of Leyton & Sherrington (1917), a figure which is not referred to at all in the text (more editorial indulgence?). This figure, in which the responses were grouped diagrammatically (sic), is reproduced here as Fig. 1B. Its simplicity conceals the wealth of data that each of their mapping experiments yielded (Fig. 1A). While the orderly mapping of different areas of the body is well substantiated, evidence for a strict somatotopy within each area (e.g. fixed non-overlapping ordering in the lateral to medial direction of thumb, then index, middle, ring and little finger movement) has been hard to find, and modern investigators (see Lemon, 1988; Schieber & Hibbard, 1993; Rathelot & Strick, 2006) have turned away from this concept to one much closer to the multiple and overlapping representation of movement described by Leyton & Sherrington (1917). This type of representation is well suited for the many and varied combination of 'fractional' movements into useful actions or, as they put it: '…that from movements of locally restricted parts, e.g. movements of a finger or of a limb-joint (movements themselves discrete and individually separable in the motor cortex), the upbuilding of larger combinations varied in character and serviceable for purposes of different and varied kind, prehensile, defensive, locomotor, mimetic, masticatory, deglutitional, orientational etc. is one of the main offices performed by the motor cortex.' We are now learning how individual motor cortex neurons combine and recombine their actions (Jackson et al. 2003). Many investigators have raised the issue of why the basic building blocks of all types of movements are represented at all in the neocortex, one of brain's 'higher' centres; surely these could be managed by subcortical centres, including the spinal cord? Leyton and Sherrington recognized, of course, that the spinal cord could generate many types of movements. However, 'It would seem that in order to preserve the possibility of being interchangeably compounded in a variety of ways, successive or simultaneous, these movements must lie, as more or less discrete and separable elements, within the grasp of the organ which has the varied compounding of them.' Beautifully put! They also recognized that the acquisition of new combinations would make an important contribution to the learning of motor skills: 'The acquirement of skilled movements, though certainly a process involving far wider areas of the cortex than the excitable zone itself, may be presumed to find in the motor cortex an organ whose synthetic properties are part of the physiological basis which renders that acquirement possible.' Leyton and Sherrington's seminal paper provided a key set of observations that have guided and inspired research ever since. They were clearly impressed by the degree of recovery after large cortical lesions and recognized that some kind of compensatory changes must be at work. They also recognized the importance of finding the neural substrate of recovery. It was only much later that it became clear that plasticity within the cortex is itself an active, use-dependent process, and that this process can be harnessed to improve recovery of function. Ultimately, Leyton and Sherrington's work opened up an exciting new chapter in the physiology of the cortex, which ultimately is having far-reaching consequences for our understanding of the brain and for the rehabilitation of patients with brain injury. Photograph of C. S. Sherrington supplied by The Wellcome Library, London.