Spinal motoneuron pools
Figure 2.1. Classical view (a) and current view (b) of the manner in which Ml neurons connect to the spinal motoneuron pools and their muscles.
that focal stimulation can elicit movement of the same body part from multiple sites in a broad region. Thus, regions where stimulation elicits movements of different body segments must overlap extensively. Early studies of electrical stimulation of the cortical surface, including those of Penfield, show that within the face region, arm region, or leg region, the territories from which stimulation can elicit movement of different segments overlap considerably (Leyton and Sherrington, 1917; Penfield and Boldrey, 1937; Woolsey et al., 1952). Despite the development of more focal stimulation of the cortex over the past few decades, the spatial resolution of motor maps has not become finer. Intracortical microstimulation shows a broadly overlapping mosaic of points where stimulation elicits movements of different body segments (Sato and Tanji, 1989; Waters et al., 1990; Park et al., 2001). Moreover large cortical territories have projections that converge on single spinal motoneurons (Andersen et al., 1975). Taken together, these studies support the view that output from a broad cortical territory converges on the spinal motoneuron pool for a given muscle and that the cortical territories for different muscles overlap extensively.
Second, whereas the prior view of M1 suggested that each M1 neuron with corticospinal output influences just one muscle, the current view indicates that a given M1 neuron may influence the motoneuron pools of several muscles. Evidence that single M1 neurons provide output to multiple muscles has been obtained using spike-triggered averaging, an analysis technique that is used to detect and quantify relatively direct relationships between a cortical neuron and a spinal motoneuron pool. Spike-triggered averaging studies in awake monkeys have shown that single M1 neurons usually connect to two or more muscles (Fetz and Cheney, 1980; Buys et al., 1986) and that some single M1 neurons connect to both proximal and distal muscles (McKiernan et al., 1998). Anatomically, horseradish peroxidase staining has shown that some single corticospinal axons have terminal arbors in multiple spinal motoneuron pools (Shinoda et al., 1981). Thus, although some corticospinal neurons may influence only one muscle, many others diverge to multiple muscles.
Third, whereas the prior view of M1 suggested that neurons in one region of M1 would be active during movement of one body segment and neurons in another region of M1 would be active during movement of another body segment, the current view indicates that neurons in overlapping M1 territories will be active during movements of different body segments. Evidence supporting the current view has been found in monkeys and in humans. Recording the activity of single M1 hand area neurons in monkeys trained to flex and extend each of their fingers revealed that (1) a given neuron could be active with movement of several different fingers and (2) the region of M1 where active neurons were found during movement of a given digit was co-extensive with the region where active neurons were found during the movement of any other digit (Schieber and Hibbard, 1993). Reversible inactivation studies in similarly trained monkeys show that small amounts of muscimol injected into the hand area disrupt some finger movements more than others, but the movements that are disrupted were unrelated to the mediolateral location of the inactivation along the central sulcus (Schieber and Poliakov, 1998). Similarly, other reports of small reversible (Brochier et al., 1999) and permanent (Friel and Nudo, 1998) lesions to the M1 hand area have documented impairment in movements of body segments proximal to the hand, as well as in the hand itself. In humans, regional cerebral blood flow studies have shown that, although a small somatotopic shift for the center of activation can be identified, extensively overlapping M1 regions are activated during movements of a single finger, several fingers, or of the more proximal arm segments (Grafton et al., 1993; Sanes et al., 1995; Kleinschmidt et al., 1997; Beisteiner et al., 2001; Hlustik et al., 2001; Indovina and Sanes, 2001). Likewise, cases of small lesions to the M1 hand territory suggest that although there is a general somatotopic gradient in human M1 (Lee et al., 1998; Schieber, 1999; Kim, 2001), a strict mediolateral somatotopy with discrete regions for each finger and for the wrist, elbow, and shoulder likely does not exist. Taken together, this evidence shows that natural movements of different body parts involve activation of M1 in broad, overlapping territories.
Implications of M1 organization for recovery after brain injury
What implications do these advances in our understanding of M1 have for functional recovery and neurorehabilitation after brain injury? First, note that the current view suggests that any region of M1 normally participates in the control of many body parts. Neurologists, physiatrists, and rehabilitation professionals witness this in clinical practice daily. One never sees a cortical lesion produce weakness of just the thumb, or just the little finger, or just the upper lip. Rather, any lesion affecting M1 that is large enough to produce clinical findings produces simultaneous weakness of multiple contiguous body segments. The entire face on one side, the entire hand, or the entire foot becomes weak from a cortical lesion. The weakness may be greater in the thumb and index than in the ulnar fingers, or greater distally than proximally in an extremity, but weakness is not confined to one small body segment. This observation is customarily ascribed to the fact that disease processes do not respect physiologic boundaries. But if this were the reason, some patients should have a weak face and thumb, but strong fingers, or a weak leg and little finger, but a strong thumb. Instead, these patterns do not occur because the physiologic boundaries do not exist.
If any one region of M1 normally participates in the control of many body parts, then when part of M1 is injured, spared regions of M1 may be able to restore function, not because they have assumed entirely new functions, but because they were participating in those functions before the injury. This may explain, in part, why after small experimental lesions to the M1 hand territory in monkeys with subsequent forced use of the affected hand, intra-cortical microstimulation in the territory surrounding the lesion that had originally represented more proximal muscles, now produced contractions in hand muscles (Nudo et al., 1996a, see also Volume I, Chapter 8 on Plasticity in motor functions and Chapter 14, Plasticity after Brain Lesions). Similarly, in people with small lesions of the hand knob area of M1 (Yousry et al., 1997), recovery may be nearly complete, such that impairments in motor control can be detected only during kinematic testing of independent finger movements (Lang and Schieber, 2003). Thus, if some M1 territory is spared, part of the functional recovery seen after brain injury may arise from the inherent flexibility of the remaining M1 territory.
Redistribution of function within M1 may be capable of occurring over greater cortical distances and faster than previously thought. For example, the areas of M1 devoted to control of the fingers have been shown to enlarge when a normal subject practices a complex finger movement task (Karni et al., 1995). Enlargement of the M1 hand area over several weeks has been observed with intracortical microstimulation mapping in monkeys (Nudo et al., 1996b), and studies employing transcranial magnetic stimulation suggest that in humans, such enlargement can occur within minutes (Pascual-Leone et al., 1994). Simply repositioning the forelimb in rodents has been shown to shift the boundary between the forelimb and vibrissae (whisker) representations in Ml (Sanes et al., 1992). Permanent changes in the human Ml's neural connections with the periphery, such as those occurring with spinal cord injury or amputation, have been shown to alter the M1 map (Levy et al., 1990; Giraux et al., 2001). Likewise, temporary ischemic anesthesia of the forearm for 30min produces greater activation of proximal arm muscles by a constant transcranial magnetic stimulus compared to before the ischemic anesthetic (Brasil-Neto et al., 1993). M1 can thus reorganize extensively and promptly.
An important and rapidly developing topic in neurorehabilitation research is how this capacity for flexible reorganization of M1 might be exploited in humans with brain injury to enhance functional recovery. For example, Muellbacher and colleagues are exploring how afferent input from proximal versus distal upper extremity segments may be manipulated to increase the M1 representation of the hand, and potentially result in improved hand function in people with chronic stroke (Muellbacher et al., 2002; see Volume I, Chapter 15 on Influence of theories of plasticity on humans). While it could be many years until these novel approaches might be proven efficacious enough to be used routinely in clinical practice, we speculate that neurorehabilitation will eventually be able to make use of M1's ability to flexibly redistribute its function, by identifying the injured regions and then designing specific rehabilitative strategies that make maximal use of nearby uninjured regions.
Was this article helpful?