- Email: [email protected]
Input-output organization of the foot motor area in humans
Clinical Neurophysiology 110 (1999) 1315±1320
Input-output organization of the foot motor area in humans Katsuyuki Machii a, b, Yoshikazu Ugawa a,*, Yasuo Terao a, Ritsuko Hanajima a, Toshiaki Furubayashi a, Hitoshi Mochizuki a, Yasushi Shiio a, Hiroyuki Enomoto a, Haruo Uesugi a, Shigeki Kuzuhara b, Ichiro Kanazawa a a
Department of Neurology, Division of Neuroscience, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan b Department of Neurology, School of Medicine, Mie University 2-174, Edobashi, Tsu, Mie 514-8507, Japan Accepted 8 March 1999
Abstract Objective: A well-organized input-output relation similar to that of the monkey motor cortex has been demonstrated in the human hand motor area (Terao Y, Ugawa Y, Uesaka Y, Hanajima R, Gemba-Shimizu K, Ohki Y, Kanazawa I. Input-output organization in the hand area of the human motor cortex, Electroenceph clin Neurophysiol 1995;97:375-381). The aim of this study is to investigate the input-output organization of the human foot motor area. Methods: We studied the effect of tactile stimuli given to the toe tip on the sizes of following responses; motor evoked potentials (MEPs) elicited by transcranial magnetic or electrical stimulation (TMS or TES) over the motor cortex and magnetic stimulation at the foramen magnum level. Results: Air stimuli applied to the toe tip facilitated magnetically evoked MEPs of mainly the muscle attached to that toe, although a less prominent facilitation was also noted in muscles attached to the adjacent toes. Neither responses evoked by TES, nor those by stimulation at the foramen magnum level, were affected by air stimuli. These results suggest that the observed facilitatory effect occurs at the cortical level. Conclusion: A fairly well-organized input-output relation is present also in the foot motor area in humans, although the facilitatory effect is not so topographically restricted as is noted for the hand motor area. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Input-output relation; Foot motor cortex; Air stimulation; Cortical efferent zones; Transcranial magnetic stimulation
1. Introduction RoseÂn and Asanuma reported, using the intracortical microstimulation technique, that the primary motor cortex consists of ®ne-grained efferent zones which receive sensory information from a portion of limb in close anatomical relationship to the muscles which they project (Asanuma and RoseÂn, 1972; RoseÂn and Asanuma, 1972). Using transcranial magnetic stimulation (TMS), Terao et al. (1995) demonstrated a similar input-output relation of the human hand motor area. Sensory feedback information from the skin to the motor cortex should also be important for the motor control of leg muscles. Therefore, there must be such input-output organization in the human foot motor area, although there has been no animal data on this point. On the other hand, it is apparent that the foot muscles are less organized for ®ne and independent movement than the hand muscles, and some foot muscles rather subserve function * Corresponding author. Tel.: 1 81-3-3815-5411, ext 3783; fax: 1 813-5800-6548.
important for postural control. Compared with the hand area of the primary motor cortex, much less sensory information is available in the foot area in the primate. In the present communication, using non-invasive stimulation techniques, we investigated whether a similar input-output relation also exists in the human foot motor area. This would also serve as a measure for what extent the foot motor area is topographically organized. 2. Methods The following experiments were done with the approval of the ethics committee of the University of Tokyo. We recruited 10 healthy volunteers, 9 males and one female, aged 29±44 years. They were all from our medical staff, and neurologically normal by examination, history, laboratory studies and electroencephalographic recordings. All of them gave their informed consent prior to this study. The experiments consisted of two parts. In the ®rst part, we investigated the effect of air stimuli on motor evoked potentials (MEPs) elicited by TMS which are known to be in¯u-
1388-2457/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 1388-245 7(99)00065-6
CLINPH 98118
1316
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
enced by the cortical excitability changes. In the second part, we studied its effect on MEPs elicited by transcranial electrical stimulation (TES) or magnetic stimulation at the foramen magnum level (Ugawa et al., 1994) which are not affected by the cortical excitability changes. We also studied the effect of air stimuli on H-re¯exes. However, because Hre¯exes were evoked in foot muscles in few subjects and their sizes changed in a small range even by changing the intensity of peripheral nerve stimulation, we considered them inappropriate for studying spinal excitability changes in the present investigation. Therefore, we do not present results of H-re¯exes. 2.1. Experiment I The subjects lay comfortably on a reclining chair and relaxed their foot muscles. We paid special attention to the subject's consciousness so that the subject was completely alert throughout the experiments, because the facilitatory effect depended on the alertness of subjects. To monitor the subject's consciousness, electroencephalogram was continuously recorded from the occipital electrocles during the experiment, and one of 3 examiners looked at his/her face carefully and sometimes talked with him/her. If the subject was sleepy, we stopped the experiment and he/she was allowed to take a nap until he/she was fully awake. Under this condition, the results were relatively stable and consistent for each subject. The second examiner performed magnetic stimulation, while the remaining examiner applied air ¯ow to various parts of a subject's foot. The latter examiner took care not to touch any part of the subject's body. Air stimuli were given by an air pump (AP-115RN, Rei-sea Company, Japan) which could produce a maximal ¯ow of 0.867 l/min, and a maximal pressure of 1 kgf/cm 2. Flow of compressive air was conducted through a plastic tube connected to a nozzle 2 mm in diameter. In order to monitor EMG activities in the studied muscles, audiovisual feedback was provided to both the subject and 3 examiners by an oscilloscope connected to a loudspeaker (SRS-57, Sony, Japan) with the gain set at 100 mV/cm. Magnetic stimuli were delivered through a ®gure-8shaped coil (internal diameter 4.5 cm) with the junction of two circular coil wings was placed over the foot motor area (usually the vertex (Cz in the international 10±20 system) to 2 cm posterior to Cz). This coil was connected to a Magstim 200 magnetic stimulator (Magstim Company, UK). In obtaining responses from the right foot muscles, the coil current in the junction ¯owed rightward so that the current in the brain ¯owed leftwards for activating the foot motor area in the left hemisphere and the current was reversed for the left foot muscles. MEPs were recorded with surface cup electrodes placed over the belly and tendon of abductor hallucis (AH) and abductor digiti minimi (ADM) muscles. EMG activities were ampli®ed with ®lters set at 100 Hz and 3 kHz, then recorded by a computer (Signal Processor DP-
1200, NEC Medical Systems), with which we could perform a randomized conditional averaging. Seven subjects were asked to keep these muscles relaxed. In the other 3 subjects in whom no MEPs were elicited in relaxed muscles, we asked the subjects to isotonically contract the target muscle (10% of maximal voluntary contraction). Since the results for active and relaxed muscles were essentially the same, as will be stated later in the results, we treated all the results from relaxed and active muscles as one group in statistical analyses. Randomized conditioning-test paradigm was used, in which control and conditioned trials were tested in a randomized order until we collected 8±15 responses for conditioned trials and at least 15 responses for control trials. In control trials, TMS was given alone. In conditioned trials, TMS was given approximately 0.2±0.3 s after the presentation of air ¯ow. We used this interval because we got a clear facilitatory effect on the hand motor area (Terao et al., 1995) using the same interval. We could not study the precise timing of our effect because we had no methods to regulate the timing of air ¯ow precisely. However, in a practical point of view, this method was good enough to always elicit our facilitatory effect. Actually, we have gotten reasonable results on the hand motor area in neurological patients using this interval of air stimulation (not published) and also reliable results in the experiments for the foot motor area (see below). The air ¯ow was removed from toes every time after each TMS to prevent possible habituation. The air stimuli were adjusted to the maximal ¯ow of the pump. The air ¯ow, when properly applied to the volar aspect of the toe tips, formed a slight dimple on the skin. At least 3 different intensities of cortical stimulation (usually more than 5 intensities and sometimes more than 10) were selected so that, given alone, they evoked control MEPs of around 0.1 mV, 0.5 mV and 1 mV. In a certain number of the subjects, we could not evoke responses of about 1 mV because of high threshold. The mean peak-to-peak sizes of responses of each muscle were calculated. In order to see in which condition signi®cant effect of air stimulation was present, we compared their sizes using one factor (main effect of position of the air stimulation) ANOVA test with Bonferroni corrected post hoc analysis in AH and ADM separately. The average amplitude under each condition from all the subjects was expressed as its ratio to the average amplitude of control response in order to overcome intersubject variability of sizes of control EMG responses. 2.2. Experiment II We compared the effects of air stimulus on MEPs evoked by 3 different forms of stimulation: magnetic and electrical cortical stimulation, and magnetic foramen magnum level stimulation (Ugawa et al., 1994). Electrical cortical stimulation was performed using a Digitimer D 180 stimulator (Digitimer Ltd., UK) with the anode placed over the vertex and the cathode at 6 cm lateral to it. The foramen magnum
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
level stimulation was performed with a double-cone coil placed over the inion. In this experiment, stimulation was performed during voluntary contraction of the target muscle because we intended to know the effect of air stimuli on the ®rst recruited descending volley in each method of stimulation (D-wave in TES and I-wave in TMS). Since both Dand I-waves contribute to MEPs elicited by TES when the subject makes no contraction, we can not see the effect of air stimulation on D-waves purely if recording EMG responses from relaxed muscles. In contrast, when recordings were made from active muscle, D-wave activation by TES directly re¯ects in EMG potentials. Therefore, the effect of air stimulation on D-waves can be evaluated by recording EMG potentials from active muscles. We adjusted the intensity of stimuli so that the 3 forms of stimulation elicited similar sized control responses. Trials for 6 different conditions (the conditions with and without air stimuli for each of the 3 forms of stimulation) were intermixed in randomized order in one session. The size ratios of conditioned to control responses for these 3 stimulation methods were compared to one another. We used a 0.2±0.3 s conditioning ± test interval in every form of stimulation. In our experience, the duration of air ¯ow was important for the effect on magnetic cortical responses shown below. Therefore, the interval should be 0.2±0.3 s (the duration of air ¯ow is 0.2±0.3 s). If so, the latency differences in responses to different forms of stimulation (at most several ms) must not be so critical for our effect. Based on these, we considered that the 0.2±0.3 s interval was good enough for studying the effect on responses elicited by every stimulation methods. We compared the sizes of conditioned responses with those of control responses by using Student's paired t test.
3. Results 3.1. Experiment I 3.1.1. Facilitatory effect of air stimuli on magnetically evoked MEPs In all subjects, air stimuli applied to a certain part of the ipsilateral foot had a facilitatory effect on magnetically evoked MEPs either when the studied muscle was active or relaxed. A typical response pattern obtained in one subject is shown in Fig. 1. The peak-to-peak MEP amplitude of AH was markedly enlarged by air applied to the volar surface of the great toe and moderately by air to the second toe, but not by those to the other toes. On the other hand, the MEP of ADM was enlarged when air was applied to all the toes except the great toe. The effect was prominent when air was applied to the 5th toe, and was progressively smaller when air was applied to toes farther apart from the recorded muscle. Thus, the MEP of a muscle attached to one toe was facilitated by air stimuli applied to the tip of that toe and,
1317
Fig. 1. A typical pattern of facilitatory effect of air stimuli on MEPs evoked by magnetic cortical stimulation. Responses from AH are shown on the left, and those from ADM on the right. Top traces are control responses and the other responses conditioned by air stimuli. Numbers on the left indicate the toe numbers to which air was applied. The MEP of AH was markedly enlarged when air was applied to the great toe, and moderately so when applied to the second toe, but no changes when applied to the other toes. Responses of ADM were greatly enlarged by air stimuli applied to the 5th toe, moderately when applied to the 4th toe, and slightly when applied to the second and 3rd toes. Air stimuli on the great toe had no effect.
sometimes, to the neighboring toes, whereas facilitation was less apparent when air was applied to toes farther apart. Similar dependence of our effect on the site of air stimulation was observed in all the subjects. Results from the sessions in which control responses were smaller than 0.3 mV were used for studying the correlation between the site of air stimulation and facilitatory effects because our effect depended on the size of control responses (see below). The means and standard errors of the size ratios (conditioned/ control) obtained from these results were calculated across all the subjects at each stimulation site (Fig. 2). In AH, a signi®cant facilitation of MEP was observed with stimuli on the great and second toes (P , 0:05, ANOVA; post hoc analysis). Air stimuli on the 3rd toe tended to facilitate MEPs, but their effect was not statistically signi®cant, probably because of the small number of subjects we studied. On the other hand, in AH the MEP was facilitated by air applied to the 3rd, 4th and 5th toes (P , 0:05, ANOVA; post hoc analysis). These effects did not signi®cantly (P . 0:05, post hoc analysis) differ among the 3rd to 5th toes. Air stimuli on the second toe had a tendency to facilitate MEPs of ADM, but its effect was not statistically signi®cant. In short, this facilitatory effect tended to increase when the air was applied to a toe in close proximity to the toe attached by the target muscle, although the extent of this facilitatory effect was not so topographically restricted as shown in the hand muscles (Terao et al., 1995). 3.1.2. Facilitatory effect of air stimuli on MEPs of various control sizes In Fig. 3, the ratios of conditioned to control response elicited by TMS with different intensities (different control sizes) were plotted against the sizes of control MEPs for 5
1318
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
subjects. Here facilitation was observed in all of these subjects, when the control size was less than 0.5 mV. However, the facilitatory effect became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV. 3.2. Experiment II 3.2.1. Comparisons among the effects on MEPs evoked by magnetic stimulation over the motor cortex and at the foramen magnum level and TES Fig. 4 shows differences in the facilitatory effect among responses of AH evoked by 3 forms of stimulation (TMS, TES and magnetic foramen magnum level stimulation) in a single subject. Air stimulation had a marked facilitatory effect on magnetically evoked MEPs. In contrast, neither electrically evoked MEPs nor MEPs elicited by magnetic stimulation at the foramen magnum level were enlarged by air stimuli. The mean size ratios from all the subjects in the 3 forms of stimulation are shown in Fig. 5, which con®rmed that only MEPs to TMS were facilitated by air stimuli (P , 0:05, Student's paired t test).
Fig. 2. Topographical pattern of facilitatory effect of air stimuli. Mean (^SE) size ratios obtained from all the subjects are plotted against the site of air stimulation. In the case of AH, signi®cant facilitation of MEP was observed when air stimuli were given to the great and second toes (P , 0:05, ANOVA; post hoc analysis). MEPs of ADM were signi®cantly facilitated by the stimulus applied to the 3rd, 4th and 5th toes (P , 0:05, ANOVA; post hoc analysis).
Fig. 3. The amount of facilitation and the size of control responses. The size ratios for different stimulation conditions (different intensities of stimulation) in 5 subjects were plotted against the size of control EMG responses of AH. Moderate facilitation was observed for control responses smaller than 0.5 mV, but the facilitation became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV.
Fig. 4. Facilitatory effects on 3 different responses. In each column, the left traces are control responses, and the right traces are responses conditioned by air stimulation. All responses are average responses (n 15) recorded from AH in a subject. Responses to 3 kinds of stimulation (magnetic cortical stimulation (TMS), electrical cortical stimulation (TES) and foramen magnum level stimulation (FM)) are shown. Air stimulation had a marked facilitatory effect on the magnetically evoked MEP. None of the other responses were facilitated by air stimuli.
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
Fig. 5. The mean (^SE) size ratios for all the subjects were plotted against the different kinds of control responses. Responses to TMS were signi®cantly facilitated by air stimuli (P , 0:05, Student's paired t test), but the other two responses were not.
4. Discussion Our results showed that the sensory input to the toes produced by air stimulation had a facilitatory effect on the size of magnetically evoked MEPs of foot muscles. The effect was similar to that noted for the hand motor area (Terao et al., 1995), but the peripheral sensory area whose stimulation facilitated MEPs of foot muscles was not so restricted as that for hand muscles. The facilitation was observed only for magnetically induced MEPs, while neither MEPs elicited by TES nor those by foramen magnum level stimulation were affected. TMS is considered to activate the corticospinal neurons trans-synaptically at the cortical level, whereas both TES and magnetic stimulation at the foramen magnum activate their axons directly (Rothwell et al., 1991; Ugawa et al., 1994). Then, if the change of excitability occurs in the motor cortex, responses to TMS are affected and their sizes must be changed, whereas the sizes of responses to TES or foramen magnum level stimulation are not affected because activation occurs at the corticospinal axons. Therefore, we presume that the facilitatory effect observed here is produced at the cortical level, that is, excitability of the motor cortex is altered by the sensory input. It was reported that spinal excitability is changed by afferent inputs (Gassel and Ott, 1970; Deuschl et al., 1995). The reasons why our air stimuli had no effect on spinal excitability must be as follows. Gassel et al. observed spinal facilitation mediated by Group II ®bers (probably mediating input by air stimuli) at short intervals (40±90 ms) in a leg muscle, and Deuschl et al. also observed spinal facilitation at short intervals (30±40 ms) and probably cortical facilitation at long intervals (80±100 ms) in hand muscles. Our interval of 0.2±0.3 s (200±300 ms) is much longer than the intervals for spinal facilitation in these reports. Moreover, Gassel et al. (1970) used electrical stimulation, which is much more powerful than air stimulation. This difference in the number of activated ®bers may be another factor to explain the lack of spinal excitability changes in our method. Based on this, we conclude that the
1319
input-output organization noted in the foot motor area is essentially the same as that described for the hand motor area in primates by RoseÂn and Asanuma (1972) and in humans by Terao et al. (1995). For the leg motor area, however, the topographical organization of facilitatory effect was not so restricted as for the hand motor area. Air stimulation had a facilitatory effect not only on the MEPs of muscles attached to that toe to which air was applied, but also on MEPs of muscles attached to the adjacent toes. In the following, we make some speculation from the previous studies as to the difference between hand and foot muscles. At ®rst, certain foot muscles may subserve motor functions distinct from those of the hand muscles. Many foot muscles are involved in the maintenance of posture while the primary role of hand muscles resides in the ®ner manipulation of objects. As a result, the movement of the ®ngers, especially in humans, may be more delicately organized than those of the toes. This is re¯ected in the larger and probably ®ner organization of the cortical representation of the hand muscles than those innervating the leg muscles. In monkeys, the representations of individual leg muscles show much overlap (Jankowska et al., 1975). There was evidence that adjacent pyramidal tract neurons in the leg area of the motor cortex may make synaptic contacts on the spinal motoneuron pool innervating the same muscle (convergence) and, in the other way round, branches of the same pyramidal tract neuron may project to several motoneuron pools (divergence) (Asanuma et al., 1979). This lenience of innervation should be greater in the leg motor area than in the hand motor area, which must obscure the input-output relation in the cortex. Brouwer and Ashby (1992) examined the projections of cortical neurons activated by TMS to a single lower limb spinal motoneuron in man. They constructed peristimulus time histograms (PSTHs) of motor unit ®ring after the magnetic stimulation. Motor units recorded from the tibialis anterior (TA) muscle showed the strong short latency facilitation, but not in the other muscles. For soleus, for example, many units exhibited short latency inhibition without preceding facilitation. This result was consistent with those of Cowan et al. (1986) who suggested that a monosynaptic excitatory cortical projection, which is presumed to be responsible for the large facilitation of the H-re¯ex in muscles of the upper limb and TA muscles, is small or absent in the soleus muscle. According to the mode of facilitation evoked by TMS (monosynaptic facilitatory connection), spinal motoneurons innervating the lower limb muscles would be classi®ed between those receiving strong monosynaptic projections by the corticospinal tract (such as TA) and polysynaptic projections from the cortex (predominant inhibition, such as soleus). AH or ADM, which we used in this study, must have moderate amount of monosynaptic excitatory cortical projections. These imply that as compared with the motoneuron pools of hand muscles, those innervating the leg muscles may contain a greater proportion of motor units receiving polysynaptic inputs from the
1320
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
cortex and subcortical structures. Since the motoneurons for hand muscles receive more monosynaptic inputs from the cortex, the input-output organization of the motor cortex may be more directly re¯ected in the hand muscles than in the leg muscles. The above could be another factor which may make the input-output organization less precise in the leg motor area than in the hand motor area in our study using magnetic stimulation. The facilitatory effect was dependent on the control size. When control responses were smaller than 0.5 mV, a clear facilitation was observed, but the facilitation became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV. Similar dependence of facilitatory effects on the size of control response has been repeatedly reported for MEPs from the hand muscles (Maertens de Noordhout et al., 1992; Ohki et al., 1994; Terao et al., 1995). The mechanism for this control size dependency of several modulatory effects on motor cortical excitability remains to be determined. One possible explanation is that the corticospinal neurons early recruited by magnetic stimulation (contributing small MEPs) are ®nely regulated by several modulatory inputs, whereas those recruited later (contributing only large MEPs) are resistant to such modulators. If so, small responses mostly produced by activation of early recruited corticospinal neurons are clearly facilitated by air stimuli, but large responses produced by a large amount of discharges of lately recruited corticospinal neurons are not so much facilitated. Despite unclear mechanisms, this dependency is a kind of universal phenomenon in the modulation of MEP sizes. Even though we can not have a ®rm conclusion because of a small number of subjects we studied, we propose the following conclusions based on the present results. We have shown that, in the foot motor area, there is a input-output relation similar to that of the hand motor area. This interaction is considered to occur at the cortical level, i.e. the motor cortical excitability is altered by the sensory input. However, this input-output relation of the human foot motor area is not so topographically restricted as that of
the hand motor area. This could re¯ect, as compared with hand, smaller cortical representations for the leg muscles and the difference in function and organization of motor cortical neurons for the hand and leg muscles.
References Asanuma H, RoseÂn I. Topographical organization of cortical efferent zones projecting to distal forelimb muscles in the monkey. Exp Brain Res 1972;14:243±256. Asanuma H, Zarzecki P, Jankowska E, Hongo H, Marcus S. Projection of individual pyramidal tract neurons to lumbar motor nuclei of the monkey. Exp Brain Res 1979;34:73±89. Brouwer B, Ashby P. Corticospinal projections to lower limb motoneurons in man. Exp Brain Res 1992;89:649±654. Cowan JMA, Day BL, Marsden C, Rothwell JC. The effect of percutaneous motor cortex stimulation on H re¯exes in muscles of the arm and leg in intact man. J Physiol (Lond) 1986;377:333±347. Deuschl G, Feifel E, Guschbauer B, LuÈcking CH. Hand muscle re¯exes following air puff stimulation. Exp Brain Res 1995;105:138±146. Gassel MM, Ott KH. Local sign and late effects on motoneuron excitability of cutaneous stimulation in man. Brain 1970;93:95±106. Jankowska E, Padel Y, Tanaka R. Projections of pyramidal tract cells to (motoneurones innervating hind-limb in the monkey. J Physiol (Lond) 1975;249:637±667. Maertens de Noordhout A, Rothwell JC, Day BL, Dressler D, Nakashima K, Thompson PD, Marsden CD. Effect of digital nerve stimuli on responses to electrical or magnetic stimulation of the human brain. J Physiol (Lond) 1992;447:535±548. Ohki Y, Suzuki T, Ugawa Y, Uesaka Y, Sakai K, Kanazawa I. Excitation of the motor cortex associated with the E2 phase of cutaneous re¯exes in man. Brain Res 1994;633:343±347. RoseÂn I, Asanuma H. Peripheral afferent inputs to the forelimb area of the monkey motor cortex: input-output relations. Exp Brain Res 1972;14:257±273. Rothwell JC, Thompson PD, Day BL, Boyd S, Marsden CD. Stimulation of the human motor cortex through the scalp. Exp Physiol 1991;76:159± 200. Terao Y, Ugawa Y, Uesaka Y, Hanajima R, Gemba-Shimizu K, Ohki Y, Kanazawa I. Input-output organization in the hand area of the human motor cortex. Electroenceph clin Neurophysiol 1995;97:375±381. Ugawa Y, Uesaka Y, Terao Y, Hanajima R, Kanazawa I. Magnetic stimulation of corticospinal pathways at the foramen magnum level in humans. Ann Neurol 1994;36:618±624.
Input-output organization of the foot motor area in humans Katsuyuki Machii a, b, Yoshikazu Ugawa a,*, Yasuo Terao a, Ritsuko Hanajima a, Toshiaki Furubayashi a, Hitoshi Mochizuki a, Yasushi Shiio a, Hiroyuki Enomoto a, Haruo Uesugi a, Shigeki Kuzuhara b, Ichiro Kanazawa a a
Department of Neurology, Division of Neuroscience, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan b Department of Neurology, School of Medicine, Mie University 2-174, Edobashi, Tsu, Mie 514-8507, Japan Accepted 8 March 1999
Abstract Objective: A well-organized input-output relation similar to that of the monkey motor cortex has been demonstrated in the human hand motor area (Terao Y, Ugawa Y, Uesaka Y, Hanajima R, Gemba-Shimizu K, Ohki Y, Kanazawa I. Input-output organization in the hand area of the human motor cortex, Electroenceph clin Neurophysiol 1995;97:375-381). The aim of this study is to investigate the input-output organization of the human foot motor area. Methods: We studied the effect of tactile stimuli given to the toe tip on the sizes of following responses; motor evoked potentials (MEPs) elicited by transcranial magnetic or electrical stimulation (TMS or TES) over the motor cortex and magnetic stimulation at the foramen magnum level. Results: Air stimuli applied to the toe tip facilitated magnetically evoked MEPs of mainly the muscle attached to that toe, although a less prominent facilitation was also noted in muscles attached to the adjacent toes. Neither responses evoked by TES, nor those by stimulation at the foramen magnum level, were affected by air stimuli. These results suggest that the observed facilitatory effect occurs at the cortical level. Conclusion: A fairly well-organized input-output relation is present also in the foot motor area in humans, although the facilitatory effect is not so topographically restricted as is noted for the hand motor area. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Input-output relation; Foot motor cortex; Air stimulation; Cortical efferent zones; Transcranial magnetic stimulation
1. Introduction RoseÂn and Asanuma reported, using the intracortical microstimulation technique, that the primary motor cortex consists of ®ne-grained efferent zones which receive sensory information from a portion of limb in close anatomical relationship to the muscles which they project (Asanuma and RoseÂn, 1972; RoseÂn and Asanuma, 1972). Using transcranial magnetic stimulation (TMS), Terao et al. (1995) demonstrated a similar input-output relation of the human hand motor area. Sensory feedback information from the skin to the motor cortex should also be important for the motor control of leg muscles. Therefore, there must be such input-output organization in the human foot motor area, although there has been no animal data on this point. On the other hand, it is apparent that the foot muscles are less organized for ®ne and independent movement than the hand muscles, and some foot muscles rather subserve function * Corresponding author. Tel.: 1 81-3-3815-5411, ext 3783; fax: 1 813-5800-6548.
important for postural control. Compared with the hand area of the primary motor cortex, much less sensory information is available in the foot area in the primate. In the present communication, using non-invasive stimulation techniques, we investigated whether a similar input-output relation also exists in the human foot motor area. This would also serve as a measure for what extent the foot motor area is topographically organized. 2. Methods The following experiments were done with the approval of the ethics committee of the University of Tokyo. We recruited 10 healthy volunteers, 9 males and one female, aged 29±44 years. They were all from our medical staff, and neurologically normal by examination, history, laboratory studies and electroencephalographic recordings. All of them gave their informed consent prior to this study. The experiments consisted of two parts. In the ®rst part, we investigated the effect of air stimuli on motor evoked potentials (MEPs) elicited by TMS which are known to be in¯u-
1388-2457/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 1388-245 7(99)00065-6
CLINPH 98118
1316
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
enced by the cortical excitability changes. In the second part, we studied its effect on MEPs elicited by transcranial electrical stimulation (TES) or magnetic stimulation at the foramen magnum level (Ugawa et al., 1994) which are not affected by the cortical excitability changes. We also studied the effect of air stimuli on H-re¯exes. However, because Hre¯exes were evoked in foot muscles in few subjects and their sizes changed in a small range even by changing the intensity of peripheral nerve stimulation, we considered them inappropriate for studying spinal excitability changes in the present investigation. Therefore, we do not present results of H-re¯exes. 2.1. Experiment I The subjects lay comfortably on a reclining chair and relaxed their foot muscles. We paid special attention to the subject's consciousness so that the subject was completely alert throughout the experiments, because the facilitatory effect depended on the alertness of subjects. To monitor the subject's consciousness, electroencephalogram was continuously recorded from the occipital electrocles during the experiment, and one of 3 examiners looked at his/her face carefully and sometimes talked with him/her. If the subject was sleepy, we stopped the experiment and he/she was allowed to take a nap until he/she was fully awake. Under this condition, the results were relatively stable and consistent for each subject. The second examiner performed magnetic stimulation, while the remaining examiner applied air ¯ow to various parts of a subject's foot. The latter examiner took care not to touch any part of the subject's body. Air stimuli were given by an air pump (AP-115RN, Rei-sea Company, Japan) which could produce a maximal ¯ow of 0.867 l/min, and a maximal pressure of 1 kgf/cm 2. Flow of compressive air was conducted through a plastic tube connected to a nozzle 2 mm in diameter. In order to monitor EMG activities in the studied muscles, audiovisual feedback was provided to both the subject and 3 examiners by an oscilloscope connected to a loudspeaker (SRS-57, Sony, Japan) with the gain set at 100 mV/cm. Magnetic stimuli were delivered through a ®gure-8shaped coil (internal diameter 4.5 cm) with the junction of two circular coil wings was placed over the foot motor area (usually the vertex (Cz in the international 10±20 system) to 2 cm posterior to Cz). This coil was connected to a Magstim 200 magnetic stimulator (Magstim Company, UK). In obtaining responses from the right foot muscles, the coil current in the junction ¯owed rightward so that the current in the brain ¯owed leftwards for activating the foot motor area in the left hemisphere and the current was reversed for the left foot muscles. MEPs were recorded with surface cup electrodes placed over the belly and tendon of abductor hallucis (AH) and abductor digiti minimi (ADM) muscles. EMG activities were ampli®ed with ®lters set at 100 Hz and 3 kHz, then recorded by a computer (Signal Processor DP-
1200, NEC Medical Systems), with which we could perform a randomized conditional averaging. Seven subjects were asked to keep these muscles relaxed. In the other 3 subjects in whom no MEPs were elicited in relaxed muscles, we asked the subjects to isotonically contract the target muscle (10% of maximal voluntary contraction). Since the results for active and relaxed muscles were essentially the same, as will be stated later in the results, we treated all the results from relaxed and active muscles as one group in statistical analyses. Randomized conditioning-test paradigm was used, in which control and conditioned trials were tested in a randomized order until we collected 8±15 responses for conditioned trials and at least 15 responses for control trials. In control trials, TMS was given alone. In conditioned trials, TMS was given approximately 0.2±0.3 s after the presentation of air ¯ow. We used this interval because we got a clear facilitatory effect on the hand motor area (Terao et al., 1995) using the same interval. We could not study the precise timing of our effect because we had no methods to regulate the timing of air ¯ow precisely. However, in a practical point of view, this method was good enough to always elicit our facilitatory effect. Actually, we have gotten reasonable results on the hand motor area in neurological patients using this interval of air stimulation (not published) and also reliable results in the experiments for the foot motor area (see below). The air ¯ow was removed from toes every time after each TMS to prevent possible habituation. The air stimuli were adjusted to the maximal ¯ow of the pump. The air ¯ow, when properly applied to the volar aspect of the toe tips, formed a slight dimple on the skin. At least 3 different intensities of cortical stimulation (usually more than 5 intensities and sometimes more than 10) were selected so that, given alone, they evoked control MEPs of around 0.1 mV, 0.5 mV and 1 mV. In a certain number of the subjects, we could not evoke responses of about 1 mV because of high threshold. The mean peak-to-peak sizes of responses of each muscle were calculated. In order to see in which condition signi®cant effect of air stimulation was present, we compared their sizes using one factor (main effect of position of the air stimulation) ANOVA test with Bonferroni corrected post hoc analysis in AH and ADM separately. The average amplitude under each condition from all the subjects was expressed as its ratio to the average amplitude of control response in order to overcome intersubject variability of sizes of control EMG responses. 2.2. Experiment II We compared the effects of air stimulus on MEPs evoked by 3 different forms of stimulation: magnetic and electrical cortical stimulation, and magnetic foramen magnum level stimulation (Ugawa et al., 1994). Electrical cortical stimulation was performed using a Digitimer D 180 stimulator (Digitimer Ltd., UK) with the anode placed over the vertex and the cathode at 6 cm lateral to it. The foramen magnum
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
level stimulation was performed with a double-cone coil placed over the inion. In this experiment, stimulation was performed during voluntary contraction of the target muscle because we intended to know the effect of air stimuli on the ®rst recruited descending volley in each method of stimulation (D-wave in TES and I-wave in TMS). Since both Dand I-waves contribute to MEPs elicited by TES when the subject makes no contraction, we can not see the effect of air stimulation on D-waves purely if recording EMG responses from relaxed muscles. In contrast, when recordings were made from active muscle, D-wave activation by TES directly re¯ects in EMG potentials. Therefore, the effect of air stimulation on D-waves can be evaluated by recording EMG potentials from active muscles. We adjusted the intensity of stimuli so that the 3 forms of stimulation elicited similar sized control responses. Trials for 6 different conditions (the conditions with and without air stimuli for each of the 3 forms of stimulation) were intermixed in randomized order in one session. The size ratios of conditioned to control responses for these 3 stimulation methods were compared to one another. We used a 0.2±0.3 s conditioning ± test interval in every form of stimulation. In our experience, the duration of air ¯ow was important for the effect on magnetic cortical responses shown below. Therefore, the interval should be 0.2±0.3 s (the duration of air ¯ow is 0.2±0.3 s). If so, the latency differences in responses to different forms of stimulation (at most several ms) must not be so critical for our effect. Based on these, we considered that the 0.2±0.3 s interval was good enough for studying the effect on responses elicited by every stimulation methods. We compared the sizes of conditioned responses with those of control responses by using Student's paired t test.
3. Results 3.1. Experiment I 3.1.1. Facilitatory effect of air stimuli on magnetically evoked MEPs In all subjects, air stimuli applied to a certain part of the ipsilateral foot had a facilitatory effect on magnetically evoked MEPs either when the studied muscle was active or relaxed. A typical response pattern obtained in one subject is shown in Fig. 1. The peak-to-peak MEP amplitude of AH was markedly enlarged by air applied to the volar surface of the great toe and moderately by air to the second toe, but not by those to the other toes. On the other hand, the MEP of ADM was enlarged when air was applied to all the toes except the great toe. The effect was prominent when air was applied to the 5th toe, and was progressively smaller when air was applied to toes farther apart from the recorded muscle. Thus, the MEP of a muscle attached to one toe was facilitated by air stimuli applied to the tip of that toe and,
1317
Fig. 1. A typical pattern of facilitatory effect of air stimuli on MEPs evoked by magnetic cortical stimulation. Responses from AH are shown on the left, and those from ADM on the right. Top traces are control responses and the other responses conditioned by air stimuli. Numbers on the left indicate the toe numbers to which air was applied. The MEP of AH was markedly enlarged when air was applied to the great toe, and moderately so when applied to the second toe, but no changes when applied to the other toes. Responses of ADM were greatly enlarged by air stimuli applied to the 5th toe, moderately when applied to the 4th toe, and slightly when applied to the second and 3rd toes. Air stimuli on the great toe had no effect.
sometimes, to the neighboring toes, whereas facilitation was less apparent when air was applied to toes farther apart. Similar dependence of our effect on the site of air stimulation was observed in all the subjects. Results from the sessions in which control responses were smaller than 0.3 mV were used for studying the correlation between the site of air stimulation and facilitatory effects because our effect depended on the size of control responses (see below). The means and standard errors of the size ratios (conditioned/ control) obtained from these results were calculated across all the subjects at each stimulation site (Fig. 2). In AH, a signi®cant facilitation of MEP was observed with stimuli on the great and second toes (P , 0:05, ANOVA; post hoc analysis). Air stimuli on the 3rd toe tended to facilitate MEPs, but their effect was not statistically signi®cant, probably because of the small number of subjects we studied. On the other hand, in AH the MEP was facilitated by air applied to the 3rd, 4th and 5th toes (P , 0:05, ANOVA; post hoc analysis). These effects did not signi®cantly (P . 0:05, post hoc analysis) differ among the 3rd to 5th toes. Air stimuli on the second toe had a tendency to facilitate MEPs of ADM, but its effect was not statistically signi®cant. In short, this facilitatory effect tended to increase when the air was applied to a toe in close proximity to the toe attached by the target muscle, although the extent of this facilitatory effect was not so topographically restricted as shown in the hand muscles (Terao et al., 1995). 3.1.2. Facilitatory effect of air stimuli on MEPs of various control sizes In Fig. 3, the ratios of conditioned to control response elicited by TMS with different intensities (different control sizes) were plotted against the sizes of control MEPs for 5
1318
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
subjects. Here facilitation was observed in all of these subjects, when the control size was less than 0.5 mV. However, the facilitatory effect became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV. 3.2. Experiment II 3.2.1. Comparisons among the effects on MEPs evoked by magnetic stimulation over the motor cortex and at the foramen magnum level and TES Fig. 4 shows differences in the facilitatory effect among responses of AH evoked by 3 forms of stimulation (TMS, TES and magnetic foramen magnum level stimulation) in a single subject. Air stimulation had a marked facilitatory effect on magnetically evoked MEPs. In contrast, neither electrically evoked MEPs nor MEPs elicited by magnetic stimulation at the foramen magnum level were enlarged by air stimuli. The mean size ratios from all the subjects in the 3 forms of stimulation are shown in Fig. 5, which con®rmed that only MEPs to TMS were facilitated by air stimuli (P , 0:05, Student's paired t test).
Fig. 2. Topographical pattern of facilitatory effect of air stimuli. Mean (^SE) size ratios obtained from all the subjects are plotted against the site of air stimulation. In the case of AH, signi®cant facilitation of MEP was observed when air stimuli were given to the great and second toes (P , 0:05, ANOVA; post hoc analysis). MEPs of ADM were signi®cantly facilitated by the stimulus applied to the 3rd, 4th and 5th toes (P , 0:05, ANOVA; post hoc analysis).
Fig. 3. The amount of facilitation and the size of control responses. The size ratios for different stimulation conditions (different intensities of stimulation) in 5 subjects were plotted against the size of control EMG responses of AH. Moderate facilitation was observed for control responses smaller than 0.5 mV, but the facilitation became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV.
Fig. 4. Facilitatory effects on 3 different responses. In each column, the left traces are control responses, and the right traces are responses conditioned by air stimulation. All responses are average responses (n 15) recorded from AH in a subject. Responses to 3 kinds of stimulation (magnetic cortical stimulation (TMS), electrical cortical stimulation (TES) and foramen magnum level stimulation (FM)) are shown. Air stimulation had a marked facilitatory effect on the magnetically evoked MEP. None of the other responses were facilitated by air stimuli.
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
Fig. 5. The mean (^SE) size ratios for all the subjects were plotted against the different kinds of control responses. Responses to TMS were signi®cantly facilitated by air stimuli (P , 0:05, Student's paired t test), but the other two responses were not.
4. Discussion Our results showed that the sensory input to the toes produced by air stimulation had a facilitatory effect on the size of magnetically evoked MEPs of foot muscles. The effect was similar to that noted for the hand motor area (Terao et al., 1995), but the peripheral sensory area whose stimulation facilitated MEPs of foot muscles was not so restricted as that for hand muscles. The facilitation was observed only for magnetically induced MEPs, while neither MEPs elicited by TES nor those by foramen magnum level stimulation were affected. TMS is considered to activate the corticospinal neurons trans-synaptically at the cortical level, whereas both TES and magnetic stimulation at the foramen magnum activate their axons directly (Rothwell et al., 1991; Ugawa et al., 1994). Then, if the change of excitability occurs in the motor cortex, responses to TMS are affected and their sizes must be changed, whereas the sizes of responses to TES or foramen magnum level stimulation are not affected because activation occurs at the corticospinal axons. Therefore, we presume that the facilitatory effect observed here is produced at the cortical level, that is, excitability of the motor cortex is altered by the sensory input. It was reported that spinal excitability is changed by afferent inputs (Gassel and Ott, 1970; Deuschl et al., 1995). The reasons why our air stimuli had no effect on spinal excitability must be as follows. Gassel et al. observed spinal facilitation mediated by Group II ®bers (probably mediating input by air stimuli) at short intervals (40±90 ms) in a leg muscle, and Deuschl et al. also observed spinal facilitation at short intervals (30±40 ms) and probably cortical facilitation at long intervals (80±100 ms) in hand muscles. Our interval of 0.2±0.3 s (200±300 ms) is much longer than the intervals for spinal facilitation in these reports. Moreover, Gassel et al. (1970) used electrical stimulation, which is much more powerful than air stimulation. This difference in the number of activated ®bers may be another factor to explain the lack of spinal excitability changes in our method. Based on this, we conclude that the
1319
input-output organization noted in the foot motor area is essentially the same as that described for the hand motor area in primates by RoseÂn and Asanuma (1972) and in humans by Terao et al. (1995). For the leg motor area, however, the topographical organization of facilitatory effect was not so restricted as for the hand motor area. Air stimulation had a facilitatory effect not only on the MEPs of muscles attached to that toe to which air was applied, but also on MEPs of muscles attached to the adjacent toes. In the following, we make some speculation from the previous studies as to the difference between hand and foot muscles. At ®rst, certain foot muscles may subserve motor functions distinct from those of the hand muscles. Many foot muscles are involved in the maintenance of posture while the primary role of hand muscles resides in the ®ner manipulation of objects. As a result, the movement of the ®ngers, especially in humans, may be more delicately organized than those of the toes. This is re¯ected in the larger and probably ®ner organization of the cortical representation of the hand muscles than those innervating the leg muscles. In monkeys, the representations of individual leg muscles show much overlap (Jankowska et al., 1975). There was evidence that adjacent pyramidal tract neurons in the leg area of the motor cortex may make synaptic contacts on the spinal motoneuron pool innervating the same muscle (convergence) and, in the other way round, branches of the same pyramidal tract neuron may project to several motoneuron pools (divergence) (Asanuma et al., 1979). This lenience of innervation should be greater in the leg motor area than in the hand motor area, which must obscure the input-output relation in the cortex. Brouwer and Ashby (1992) examined the projections of cortical neurons activated by TMS to a single lower limb spinal motoneuron in man. They constructed peristimulus time histograms (PSTHs) of motor unit ®ring after the magnetic stimulation. Motor units recorded from the tibialis anterior (TA) muscle showed the strong short latency facilitation, but not in the other muscles. For soleus, for example, many units exhibited short latency inhibition without preceding facilitation. This result was consistent with those of Cowan et al. (1986) who suggested that a monosynaptic excitatory cortical projection, which is presumed to be responsible for the large facilitation of the H-re¯ex in muscles of the upper limb and TA muscles, is small or absent in the soleus muscle. According to the mode of facilitation evoked by TMS (monosynaptic facilitatory connection), spinal motoneurons innervating the lower limb muscles would be classi®ed between those receiving strong monosynaptic projections by the corticospinal tract (such as TA) and polysynaptic projections from the cortex (predominant inhibition, such as soleus). AH or ADM, which we used in this study, must have moderate amount of monosynaptic excitatory cortical projections. These imply that as compared with the motoneuron pools of hand muscles, those innervating the leg muscles may contain a greater proportion of motor units receiving polysynaptic inputs from the
1320
K. Machii et al. / Clinical Neurophysiology 110 (1999) 1315±1320
cortex and subcortical structures. Since the motoneurons for hand muscles receive more monosynaptic inputs from the cortex, the input-output organization of the motor cortex may be more directly re¯ected in the hand muscles than in the leg muscles. The above could be another factor which may make the input-output organization less precise in the leg motor area than in the hand motor area in our study using magnetic stimulation. The facilitatory effect was dependent on the control size. When control responses were smaller than 0.5 mV, a clear facilitation was observed, but the facilitation became less remarkable as the control size increased and ultimately disappeared when control responses were larger than 1.0 mV. Similar dependence of facilitatory effects on the size of control response has been repeatedly reported for MEPs from the hand muscles (Maertens de Noordhout et al., 1992; Ohki et al., 1994; Terao et al., 1995). The mechanism for this control size dependency of several modulatory effects on motor cortical excitability remains to be determined. One possible explanation is that the corticospinal neurons early recruited by magnetic stimulation (contributing small MEPs) are ®nely regulated by several modulatory inputs, whereas those recruited later (contributing only large MEPs) are resistant to such modulators. If so, small responses mostly produced by activation of early recruited corticospinal neurons are clearly facilitated by air stimuli, but large responses produced by a large amount of discharges of lately recruited corticospinal neurons are not so much facilitated. Despite unclear mechanisms, this dependency is a kind of universal phenomenon in the modulation of MEP sizes. Even though we can not have a ®rm conclusion because of a small number of subjects we studied, we propose the following conclusions based on the present results. We have shown that, in the foot motor area, there is a input-output relation similar to that of the hand motor area. This interaction is considered to occur at the cortical level, i.e. the motor cortical excitability is altered by the sensory input. However, this input-output relation of the human foot motor area is not so topographically restricted as that of
the hand motor area. This could re¯ect, as compared with hand, smaller cortical representations for the leg muscles and the difference in function and organization of motor cortical neurons for the hand and leg muscles.
References Asanuma H, RoseÂn I. Topographical organization of cortical efferent zones projecting to distal forelimb muscles in the monkey. Exp Brain Res 1972;14:243±256. Asanuma H, Zarzecki P, Jankowska E, Hongo H, Marcus S. Projection of individual pyramidal tract neurons to lumbar motor nuclei of the monkey. Exp Brain Res 1979;34:73±89. Brouwer B, Ashby P. Corticospinal projections to lower limb motoneurons in man. Exp Brain Res 1992;89:649±654. Cowan JMA, Day BL, Marsden C, Rothwell JC. The effect of percutaneous motor cortex stimulation on H re¯exes in muscles of the arm and leg in intact man. J Physiol (Lond) 1986;377:333±347. Deuschl G, Feifel E, Guschbauer B, LuÈcking CH. Hand muscle re¯exes following air puff stimulation. Exp Brain Res 1995;105:138±146. Gassel MM, Ott KH. Local sign and late effects on motoneuron excitability of cutaneous stimulation in man. Brain 1970;93:95±106. Jankowska E, Padel Y, Tanaka R. Projections of pyramidal tract cells to (motoneurones innervating hind-limb in the monkey. J Physiol (Lond) 1975;249:637±667. Maertens de Noordhout A, Rothwell JC, Day BL, Dressler D, Nakashima K, Thompson PD, Marsden CD. Effect of digital nerve stimuli on responses to electrical or magnetic stimulation of the human brain. J Physiol (Lond) 1992;447:535±548. Ohki Y, Suzuki T, Ugawa Y, Uesaka Y, Sakai K, Kanazawa I. Excitation of the motor cortex associated with the E2 phase of cutaneous re¯exes in man. Brain Res 1994;633:343±347. RoseÂn I, Asanuma H. Peripheral afferent inputs to the forelimb area of the monkey motor cortex: input-output relations. Exp Brain Res 1972;14:257±273. Rothwell JC, Thompson PD, Day BL, Boyd S, Marsden CD. Stimulation of the human motor cortex through the scalp. Exp Physiol 1991;76:159± 200. Terao Y, Ugawa Y, Uesaka Y, Hanajima R, Gemba-Shimizu K, Ohki Y, Kanazawa I. Input-output organization in the hand area of the human motor cortex. Electroenceph clin Neurophysiol 1995;97:375±381. Ugawa Y, Uesaka Y, Terao Y, Hanajima R, Kanazawa I. Magnetic stimulation of corticospinal pathways at the foramen magnum level in humans. Ann Neurol 1994;36:618±624.