529

J. Physiol. (1986), 381, pp. 529-549 With 9 text-figure8 Printed in Great Britain

SELECTIVE FACILITATION OF DIFFERENT HAND MUSCLES BY SINGLE CORTICOSPINAL NEURONES IN THE CONSCIOUS MONKEY

BY E. J. BUYS, R. N. LEMON*, G. W. H. MANTEL* AND R. B. MUIRt From the Department of Anatomy, Medical Faculty, Erasmus University Rotterdam, The Netherlands

(Received 15 November 1985) SUMMARY

1. Post-spike facilitation of e.m.g. activity by monkey motor cortex neurones has been investigated in different hand and forearm muscles. 2. Seventy-eight neurones were recorded concurrently with between five and ten different muscles. Forty-seven neurones were identified as cortico-motor by the presence of post-spike facilitation in the spike-triggered average of at least one of the tested muscles. 3. All forty-seven cortico-motor neurones showed clear increases in activity during performance of a precision grip task by the monkey, and all of them were co-activated with the sampled muscles. 4. To assess the divergence of facilitation from a single cortico-motor neurone to different muscles, spike-triggered averages were constructed with all of the concurrently recorded muscles. The number of muscles in the sample, and the number of muscles showing post-spike facilitation, were corrected by excluding any post-spike facilitation which could have arisen by cross-talk between the different pairs of e.m.g. electrodes. 5. Most cortico-motor neurones produced post-spike facilitation in a restricted number of tested muscles. The mean number of post-spike facilitation-bearing muscles per cortico-motor cell rose from 1-4 +±0O5 (S.D.) when five muscles were sampled to 20 + 1-5 when ten were sampled. On average, each cortico-motor neurone produced post-spike facilitation in 27 % of the tested muscles. Only three of fortyseven cortico-motor neurones gave post-spike facilitation in half or more of the tested muscles. 6. The distribution pattern of post-spike facilitation among the muscles sampled with a given cortico-motor neurone was not altered when the spike-triggered averages were constructed from cortico-motor cell and e.m.g. activity recorded during two different phases of the precision grip task, or during performance of a quite different, power grip, task. 7. Cortico-motor cells which produced post-spike facilitation in two or more different muscles often did so in muscles with synergistic functions. 8. It is suggested that cortico-motor neurones may contribute to relatively * Present address and address for correspondence: Department of Anatomy, University of Cambridge, Downing Street, Cambridge CB2 3DY. t Present address: Department of Anatomy, Melbourne University, Parkville, Victoria 3052, Australia. Downloaded from J Physiol (jp.physoc.org) by on December 30, 2008

E. J. BUYS AND OTHERS 530 independent finger movements by virtue of their selective facilitation of hand muscles leading to a fractionated pattern of muscle activity. INTRODUCTION

There is good anatomical, behavioural and electrophysiological evidence to suggest that the direct cortico-motoneuronal connexions between pyramidal tract cells and spinal motoneurones innervating hand and forearm muscles are essential for the execution of relatively independent finger movements. This evidence has been cited in the accompanying paper (Lemon, Mantel & Muir, 1986) as a background to the need to identify these cortico-motor (c.m.) neurones in the conscious, behaving monkey. The technique of averaging e.m.g. from hand muscles with respect to discharges of a single cortical neurone (Fetz & Cheney, 1980; Muir & Lemon, 1983) has proved a reliable and useful tool for demonstrating cortical neurones with direct cortico-motor facilitation of these muscles. Having identified a population of c.m. neurones projecting to the muscles of the hand, it is important to investigate whether their properties give any clues as to their function during the performance of relatively independent finger movements. In a previous study (Muir & Lemon, 1983), we showed that some c.m. neurones that were active during a precision grip task which required independent finger movements, were significantly less active when the monkey carried out a power grip task, in which much less fractionation of finger movement was required. One possible means by which c.m. cells could contribute to the fractionated pattern of muscular activity seen during precision grip would be a rather restricted connectivity among the different motor nuclei, such that relatively small numbers of muscles, or even a single muscle, could be facilitated by a given c.m. cell active during such a movement. Whereas there is much experimental evidence concerning the convergence of cortical projections from many different cortical 'colonies' (Phillips & Porter, 1964) onto a single motoneurone (Landgren, Phillips & Porter, 1962; Andersen, Hagan, Phillips & Powell, 1975; Jankowska, Padel & Tanaka, 1975) and the multiple representation of a single muscle in the motor cortex (Rosen & Asanuma, 1972; Lemon, Hanby & Porter, 1976; Kwan, MacKay, Murphy & Wong, 1978; Lemon, 1981), there is comparatively little information about divergence in the output of a single c.m. neurone. In the monkey, there is anatomical evidence for highly branched collaterals of axons derived from corticospinal tract neurones (Shinoda, Yokota & Futami, 1981) including those identified as making corticomotoneuronal connexions (Lawrence, Porter & Redman, 1985). These studies suggest that a single corticospinal tract neurone may contact a large number of different motoneurones, which may lie in different motor nuclei. This last possibility was confirmed electrophysiologically for cells projecting to the lumbar motor nuclei of the monkey by Asanuma, Zarzecki, Jankowska, Hongo & Marcus (1979). The spike-triggered averaging method offers a new means of defining the 'muscle field' of a given c.m. neurone (Fetz & Cheney, 1980). This paper is a study of the distribution of post-spike facilitation (p.s.f.) produced by single c.m. neurones in spike-triggered averages (s.t.a.s) of up to ten different hand and forearm e.m.g.s. recorded during the performance of a precision grip task. A short account of this work has been published previously (Lemon, Mantel & Muir, 1984).

Downloaded from J Physiol (jp.physoc.org) by on December 30, 2008

SELECTIVE CORTICO-MOTOR FACILITATION

531

METHODS

Techniques for recording, stimulation and data analysis are fully described in the accompanying paper (Lemon et al. 1986). All s.t.a.s were constructed with at least 4000 spike events recorded during all phases of the precision grip task. We performed rigorous tests to exclude any possibility of cross-talk in the equipment used; this was important since this paper considers distribution of post-spike effects in s.t.a.s of different electromyogram (e.m.g.) records. During recording sessions, all e.m.g.s were carefully monitored and any changes in amplification were noted. E.m.g.s showing evidence of movement artifacts or mains interference were excluded from the analysis. E.m.g.s were registered from forearm muscles: abductor pollicis longus (ab.p.l.), brachioradialis (br.), extensor digitorum communis (e.d.c.), flexor carpi radialis (f.c.r.), flexor carpi ulnaris (f.c.u.), flexor digitorum profundus (f.d.p.), flexor digitorum superficialis (f.d.s.), palmaris longus (p.l.) and from intrinsic hand muscles: abductor pollicis brevis (ab.p.b.), flexor pollicis brevis (f.p.b.), adductor pollicis (ad.p.), first and second dorsal interossei (1st d.i. and 2nd d.i.), first lumbrical (1st lumb.) and abductor digiti minimi (ab.d.m.). TABLE 1. Number of motor cortex neurones that were recorded concurrently with different numbers of hand and forearm muscles, and which showed post-spike facilitation (p.s.f.) in at least one of the sampled muscles Number of Number of Number of neurones showing muscles neurones p.s.f. in at sampled tested least one muscle