Corticomotorneuronal hyper-excitability in amyotrophic lateral sclerosis

Journal of the Neurological Sciences 160 (Suppl. 1) (1998) S64–S68

Corticomotorneuronal hyper-excitability in amyotrophic lateral sclerosis Andrew Eisen*, Masashi Nakajima, Markus Weber Department of Neurology, University of British Columbia and the Neuromuscular Diseases Unit, Vancouver General Hospital, Vancouver, Canada V5 Z 1 M9

Abstract We have analysed how the behaviour of a voluntarily activated motor unit changes when subjected to 100–150 threshold cortical stimuli using peristimulus time histograms (PSTHs). This is a measure of the integrity of the corticomotoneuronal core innervating a single anterior horn cell. One hundred and thirty units in 29 patients with ALS and 35 units in eight age-matched normal controls were studied. PSTHs were constructed using 1-ms bins of stimulus triggered sweeps with a total analysis time of 250 ms (50 ms before and 200 ms after the stimulus). In ALS the primary peak of the PSTH was delayed in onset and prolonged in duration. The primary peak was further analysed by finer 0.2-ms bins, which showed in ALS there were more sub-components than normally occur. Additional sub-components in the PSTH primary peak implies a hyper-excitable corticomotoneuron that fires excessively. Excitability could be glutamate induced and / or due to failure of GABA inhibitory mechanisms. Some glutamate antagonsits may be therapeutic in ALS because of their anticonvulsant or GABergic properties rather than their anti-glutamate properties. GABA B agonists might have a role in future therapeutic combined therapies for ALS.  1998 Elsevier Science B.V. All rights reserved. Keywords: Amyotrophic lateral sclerosis; Corticomotoneuron; Excitability; Peristimulus time histogram; Magnetic cortical stimulation

1. Introduction In humans, each anterior horn cell, excepting those innervating the extra-ocular muscles and bladder wall, receive mono-synaptic input from a core of corticomotoneurons. The size of the core is plastic. It declines with age [16] and varies with the complexity of the motor function performed [20,24]. There is growing in vivo evidence suggesting that the motor cortex, with its numerous corticomotneuronal cores, is hyper-excitable in amyotrophic lateral sclerosis (ALS). For example positron emission tomography using [ 18 F]-2-fluoro-2-deoxy-D-glucose (FDG) and other ligands, have demonstrated a significantly increased metabolic activity in response to contralateral hand movement as compared to controls [1,23]. A variety of electrophysiologic experiments involving excitation of the corticospinal pathways also indicate the ALS motor cortex is hyperexcitable. The intensity required to stimulate the motor cortex in ALS is frequently reduced [12,13,27], the cortical silent period, a measure of cortical inhibition, is shortened [12,34], and magnetic stimulation using a conditioning-test paradigm with a short *Corresponding author. Fax:11-604-8754668.

interstimulus interval (,4 ms) fails to inhibit the test response in ALS [37]. It is not clear how these findings relate to the type or severity of neurological deficit (upper versus lower motor neuron), or disease duration. However, they imply that in ALS either the corticomotoneurons and / or their pre-synaptic terminals are overly active and that their modulation by local circuit inhibitory interneurons is impaired. When the core of corticomotoneurons that converge on a single spinal motoneuron is activated a descending volley of impulses is generated, which if adequately summated, will depolarize the anterior cell membrane and induce an excitatory postsynaptic potential (EPSP) [3,6,8,31,33]. The size of the EPSP is a reflection of the number of corticomotoneurons activated and, as we show here, the complexity of the EPSP gives information about how frequently the cell discharges. These measurements can be determined using peristimulus time histograms (PSTHs). The PSTH identifies changes in the firing probability of a voluntarily activated motor unit [10,19]. When the activated unit is subjected to an intervening cortical stimulus, its firing changes dramatically.In normal subjects the PSTH shows a consistent early post-stimulus period of significantly increased firing of an indexed motor unit.

0022-510X / 98 / $ – see front matter  1998 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 98 )00200-7

A. Eisen et al. / Journal of the Neurological Sciences 160 ( Suppl. 1) (1998) S64 –S68

This is referred to as the primary peak [9,25]. In this study we have analysed the primary peak in detail and in ALS it is more complex than normal. There are additional subcomponents, likely the result of a hyper-excitable corticomotoneuron which fires repetitively.

2. Subjects and methods Informed consent was provided from 29 patients with ALS aged from 27 to 79 years (mean 60.4614.8 years) and eight age-matched normal subjects (four men and four women) aged from 38 to 82 years (mean 54.3615.6 years). Preliminary results from these subjects have been reported elsewhere [30]. None of the subjects were on medication that would alter cortical excitability. The patients fulfilled the El Escorial criteria for definite ALS [7], but had early stage disease, particularly with respect to the studied limb. Disease duration estimated from the onset of first symptoms, excluding fasciculation, ranged from 3 to 72 months (mean 13.7613.6 months). The majority of patients (20 out of 29) had their disease less than 13 months. The extensor digitorum communis muscle (EDC) of the less affected limb was studied for each patient. The muscles were normally strong or minimally weak (MRC.41). Postsynaptic events occurring in single spinal motoneurons were derived from changes in the firing probability of single, voluntarily activated motor units in the EDC induced by transcranial magnetic stimulation to the motor cortex. The technique has been detailed previously [15,28]. A monopolar needle electrode was used to record motor unit potentials (MUPs) from the EDC muscle. Subjects were instructed to maintain steady firing of an index motor unit and were given auditory and visual feedback of the spike discharges. In each subject, four or five different motor units were examined in sequence. Changes in the firing probability of 35 normal and 130 ALS indexed units were studied and expressed as PSTHs, using specifically written software. Discharges of the indexed single motor units were collected into 1-ms bins of stimulus-triggered sweeps. Each sweep had a total analysis time of 250 ms, 50 ms before and 200 ms after the stimulus. The primary peak was further analysed by finer 0.2-ms bins. A Dantec Magpro II magnetic stimulator delivered threshold stimuli to the contralateral motor cortex through a large, cup-shaped round coil. A series of 100–120 stimuli were randomly delivered at intervals of 1–5 s. Cortical threshold measured 52.568.9% in controls and 55.7611.9% in patients. The difference is not significant. The total number of bins in the primary peak with counts exceeding the mean pre-stimulus background by more than 2SD was used to measure the magnitude and rise time of the composite corticomotoneuronal EPSP arising in the indexed motoneuron.

S65

Table 1 The primary peak of the PSTH Control

Age (year) Threshold (%) Primary peak Latency (ms) Amplitude (mV) Rise time (ms)

ALS

Mean

SD

Mean

SD

P value

54.3 52.5

15.6 8.9

60.4 55.7

14.8 11.9

0.32 0.49

19.6 2.79 3.7

2.7 1.43 1.1

22.1 2.55 3.7

4.42 1.51 1.9

0.0014 0.40 0.33

3. Result In Table 1 are summarized the differences in the primary peak in ALS compared to normal. The mean onset latency of the primary peak in ALS was significantly longer than normal. The mean amplitude and duration of control and ALS primary peaks were not significantly different. However, as is shown in Fig. 1 the range of the amplitude and duration was widely distributed in ALS (amplitude range 0.36–6.71 mV and duration range 1–9 ms, respectively). This compared to controls values of 0.75–7.5 mV for amplitude and 2–6 ms for duration. These measurements correlated positively (see Fig. 1). Of the 130 ALS motor units studied, 21 units (16.2%) from 15 patients had a reduced EPSP indicated by a small primary peak in the PSTH (, normal mean 22SD, 0.88

Fig. 1. Plot of the primary peak amplitude against its duration. In ALS (top) there is a significantly positive correlation between these characteristics. Also the range of values in ALS is considerably greater than normally occurs (lower).

S66

A. Eisen et al. / Journal of the Neurological Sciences 160 ( Suppl. 1) (1998) S64 –S68

with EPSPs that demonstrated temporal dispersion. The frequency of sub-components occurring within the primary peak in the PSTH is shown in Fig. 3. In most normal PSTHs there were 1–2 subcomponents and only occassionally additional ones. The maximum number of subcomponents recorded in normal PSTHs was 5. This contrasted with PSTHs of ALS patients which frequently had 3 or more sub-components (see Fig. 3).

4. Discussion

Fig. 2. Examples of PSTHs in a control subject aged 56 (top) and an ALS patient aged 58 (lower). Thirty ms of collection time is shown (10–40 ms), the stimulus was delivered at time 0 ms, which is not shown. In the control subject the primary peak, commencing at about 18.5 ms consists of a single sub-component (P1). In the patient with ALS the primary peak commenced at about 24 ms. It was long in duration (9 ms) and consisted of 3 sub-components P1, P2 and P3. The cumulative sum analysis (CUSUM) is also shown which makes it easier to identify sub-components.

mV), and 20 units (15.4%) from 14 patients had a temporally dispersed EPSP (. normal mean 12SD, 5.9 ms). When the PSTHs were re-constructed with 0.2-ms bins sub-components of the primary peak were readily recognized (see Fig. 2). In ALS there were more subcomponents than normal. These were always associated

Fig. 3. Frequency-distribution histogram of the sub-components of the primary peak in the PSTH labelled P1–P9. In controls most sub-components occurred as P1 or P2 whilst in ALS the majority occurred as P2–P5.

The onset of the primary peak in the PSTH derived from forearm or hand motor units is about 20 ms. This latency is in keeping with a fast-conducting descending volley as is typical of the corticomotoneuronal projection [33]. Several abnormalities in the primary peak of the PSTH have recently been described in ALS [4,5,15,16,25,26,28–30]. The peak may be small, dispersed and prolonged in latency. The abnormalities are at least in part supraspinal in origin and may well result from disease of the corticomotoneuron or its inhibitory modulation [4,5,16,28– 30]. A small primary peak equates with a small EPSP which frequently occurs in ALS. This is the result of incomplete temporal summation of the descending volley which is then not able to raise the threshold of the anterior horn cell sufficiently for it to discharge. A threshold cortical stimulus activates many corticomotoneurons which converge upon a single anterior horn cell (the cortical core). Individual corticomotoneurons within the core depolarize a small segment of the anterior horn cell’s surface membrane inducing a small EPSP. The EPSP induced by a threshold cortical stimulus is a summated EPSP generated by all of the corticomotoneurons converging on the indexed motor unit [8,15,18,20,24,31,33]. In normal subjects a single cortical stimulus causes the corticomotoneuron to fire repetitively [2]. If the size and synchrony of the evoked descending volley induces an EPSP of sufficient voltage the anterior horn cell will discharge. This will be reflected in the PSTH as a primary peak. Repetitive firing of the corticomotoneuron can cause multiple peaks in the primary peak, which are separated by 1.8–5 ms [9,25]. Sub-components are probably analagous to the I (indirect) waves that occur following a single, anodal, electrical stimulus delivered to the exposed monkey motor cortex [22,32]. The sub-component with the shortest latency may be a D (direct) wave as is seen with electrical stimulation of the exposed cortex. But these are generated by activation of the axon hillock whereas magnetic stimulation predominantly activates the pyramidal neuron through recurrent cortical circuitry inducing I waves. The number of sub-components seen in normal PSTHs seldom exceeded three. In contrast, PSTHs in ALS patients had three or more sub-components, and the latency of the

A. Eisen et al. / Journal of the Neurological Sciences 160 ( Suppl. 1) (1998) S64 –S68

earliest was often 2–4 ms later than the earliest subcomponent of normal PSTHs. The delay in ALS could reflect slowing in the motor pathways. However, prolongation in central motor conduction time measured using surface EMG recordings is unusual in ALS [14]. A more plausible explanation for the delay in the initial subcomponent is inadequate temporal summation of the descending volley resulting in incomplete depolarization the anterior horn cell. Additional, later occurring, I wave 2 3 n volleys (1 , 1 , 1 ) would be required to raise the threshold of the anterior horn cell to a level for it to discharge. Additional sub-components in the PSTH implies that the corticomotoneuron in ALS is hyperexcitable and fires more repetitively than normal. Other studies exploring different, largely unrelated, aspects of corticomotoneuronal physiology have also suggested that the motor cortex in ALS is hyper-excitable. This makes it unlikely that the present findings are simply compensatory for loss of corticomotoneurons and we postulate they reflect a primary pathophysiological process. There are no obvious clinical correlates to the cortical hyper-excitability of ALS. However, recent evidence suggests that some fasciculations in ALS are cortically generated and the fasciculation of ALS could then be a clinical manifestation of corticomotoneuronal repetitive firing [21]. Four to five different motor units in the same muscle of each patient were examined and there was considerable variation in the number of sub-components in the PSTH indicating the abnormality is not readily correlated with the severity of clinical deficit. Variability in the number of sub-components can be most readily explained by considering the anatomy and pathology of the corticomotoneuronal core innervating the specific anterior horn cell. One corticomotoneuron within a core innervates many anterior horn cells of the same motor neuron pool (divergence). Conversely a single anterior horn cell is innervated through several different corticomotoneurons (convergence) [33]. Only intact corticomotoneurons are likely to be hyperexcitable and fire more repetitively than normal. The descending volley will be well synchronized but there will be still be additional sub-components in the PSTH because of excessive firing. When the core becomes dysfunctional the descending volley will become temporally dispersed. The EPSP will be of long duration and be composed of additional sub-components, partly because the initial volleys will not be sufficiently strong to depolarise the anterior horn cell membrane, but also because of repetitive firing. Eventually the core will be composed of dead or dying corticomotoneurons which will result in a small and simple primary peak with few or no additional sub-components. Preliminary data suggests repetitive firing, as measured by the number of sub-components in the PSTH, decreases with disease progression. Enhanced excitation or reduced inhibition could both be key factors in the development of hyper-excitability of the corticomotoneuron. The evidence for increased accumula-

S67

tion of glutamate excitatory transmitter, which could stimulate the corticomotoneuron to discharge excessively, is compelling [35]. However, equally, cortical hyper-excitability in ALS could be secondary to failure of local circuit inhibitory interneurons to properly modulate corticomotoneuronal firing. g-aminobutyric acid (GABA) is the most powerful inhibitory transmitter in the nervous system and]about 10% of all neocortical neurons are GABA-accumulating local circuit neurons [11]. GABA B receptors are even present on excitatory terminals and diffusion of GABA to adjacent glutamatergic terminals curtails the release of excitatory amino acids [36]. Several physiological measures which specifically assess the inhibitory circuits have been shown to be abnormal in ALS [17]. There may be therapeutic implications to the findings of this study. They support the notion that anti-excitant therapy is rational in ALS. It is of interest that both, riluzole and neurontin, ostensibly used for their anti-glutamate properties in ALS are may also have GABAergic properties and these might be responsible for some of their reported efficacy. Lamotrigine also has anti-glutamate and anticonvulsant properties. GABA B agonists might be a useful addition to combination therapies directed to neuroprotection in ALS. It seems that with disease progression cortical excitability decreases so that later in the course of ALS use of gutamate antagonists may be counter-productive or potentially dangerous.

References [1] Abrahams S, Goldstein LH, Kew JJM, Leigh PN, et al. Frontal lobe dysfunction in amyotrophic lateral sclerosis, a PET study. Brain 1996;119:2105–20. [2] Amassian VE, Stewart M, Quirk GJ, Rosenthal JL. Physiological basis of motor effects of a transient stimulus to cerebral cortex. Neurosurgery 1987;20:74–93. [3] Ashby P, Zilm D. Relationship between EPSP shape and crosscorrelation profile explored by computer simulation for studies on human motoneurons. Exp Brain Res 1982;47:33–40. [4] Awiszus F, Feistner H. Abnormal EPSPs evoked by magnetic brain stimulation in hand muscle motoneurons of patients with amyotrophic lateral sclerosis. Electroencephalogr Clin Neurophysiol 1993;89:408–14. [5] Awiszus F, Feistner H. Comparison of single motor unit responses to transcranial magnetic and peroneal nerve stimulation in the tibialis anterior muscle of patients with amyotrophic lateral sclerosis. Electroencephalogr Clin Neurophysiol 1995;97:90–5. [6] Bawa P, Lemon RN. Recruitment of motor units in response to transcranial magnetic stimulation in man. J Physiol (Lond) 1993;471:445–64. [7] Brooks BR. World Federation of Neurology Sub Committee on Neuromuscular Diseases. El Escorial criteria for the diagnosis of amyotrophic lateral sclerosis. J Neurol Sci 1994;124(Suppl):96–107. [8] Brouwer B, Ashby P. Corticospinal projections to upper and lower limb spinal motoneurons in man. Electroencephalogr Clin Neurophysiol 1990;76:509–19. [9] Day BL, Rothwell JC, Thompson PD, Dick JPR, Cowan JMA, Berardelli A, Marsden CD. Motor cortex stimulation in intact man II. Multiple descending volleys. Brain 1987;110:1191–209.

S68

A. Eisen et al. / Journal of the Neurological Sciences 160 ( Suppl. 1) (1998) S64 –S68

[10] Day BL, Dressler D, Maertens de Noordhout A, Marsden CD, Nakashima K, Rothwell JC, Thompson PD. Electrical and magnetic stimulation of the human motor cortex: surface EMG and single motor unit responses. J Physiol (Lond) 1989;412:449–73. [11] Deisz RA, Luhmann HJ. Development of cortical excitation and inhibition. In: Gutnick MJ, Mody I, editors. The Cortical Neuron. New York: Oxford University Press, 1995:230–46. [12] Desiato MT, Caramia MD. Towards a neurophysiological marker of amyotrophic lateral sclerosis as revealed by changes in cortical excitability. Electroencephalogr Clin Neurophysiol 1997;105:1–7. [13] Eisen A, Pant B, Stewart H. Cortical excitability in amyotrophic lateral sclerosis: a clue to pathogenesis. Can J Neurol Sci 1993;20:11–6. [14] Eisen A. Amyotrophic lateral sclerosis is a multifactorial disease. Muscle Nerve 1995;18:741–52. [15] Eisen A, Entezari-Taher M, Stewart H. Cortical projections to spinal motoneurons: changes with aging and amyotrophic lateral sclerosis. Neurology 1996;46:1396–404. [16] Eisen A, Nakajima M, Enterzari-Taher M, Stewart H. Corticomotoneuron: aging, sporadic amyotrophic lateral sclerosis (ALS) and first degree relatives. In: Kimura J, Kaji R, editors. Physiology of ALS and Related Diseases. Holland: Elsevier Science, 1997:155– 75. [17] Eisen A., Nakajima M., Weber M. The motor cortex in ALS is hyper-excitable. Electroencephalogr Clin Neurophysiol (in press). [18] Fetz EE, Cheney PD. Post-spike facilitation of forelimb muscle activity by primate corticomotoneuronal cells. J Neurophysiol 1980;44:751–72. [19] Fetz EE, Gustafsson B. Relation between shapes of post-synaptic potentials and changes in firing probability of cat motoneurons. J Physiol (Lond) 1983;341:387–410. [20] Flament D, Goldsmith P, Buckley JC, Lemon RN. Task-dependence of responses in first dorsal interosseus muscle to magnetic brain stimulation in man. J Physiol 1993;464:361–78. [21] Kaji R, Kohara N, Kimura J. Fasciculations evoked by magnetic cortical stimulation in patients with ALS. Neurology 1993;43:A257–8. [22] Kernell D, Chen-Ping Wu. Responses of pyramidal tract to stimulation of the baboon’s motor cortex. J Physiol (Lond) 1967;191:653– 72. [23] Kew JJ, Brooks DJ, Passingham RE, Rothwell JC, Frackowiak BS, Leigh PN. Cortical function in progressive lower motor neuron disorders and amyotrophic lateral sclerosis: a comparative PET study. Neurology 1994;44:1101–10.

[24] Lemon RN, Mantel GWH, Muir RB. Corticospinal facilitation of hand muscles during voluntary movement in the conscious monkey. J Physiol (Lond) 1986;381:497–527. [25] Mills KR. Motor neuron disease. studies of the corticospinal excitation of single motor neurons by magnetic brain stimulation. Brain 1995;118:971–82. [26] Mills KR, Kohara N. Magnetic brain stimulation in ALS: single motor unit studies. In: Kimura J, Kaji R, editors. Physiology of ALS and Related Diseases. Amsterdam: Elsevier Science B.V., 1997:177– 92. [27] Mills KR, Nithi KA. Corticomotor threshold is reduced in early sporadic amyotrophic lateral sclerosis. Muscle Nerve 1997;20:1137– 41. [28] Nakajima M, Eisen A, McCarthy R, Olney RK, Aminoff MJ. Reduced corticomotoneuronal excitatory postsynaptic potentials (EPSPs) with normal Ia afferent EPSPs in amyotrophic lateral sclerosis. Neurology 1996;47:1555–61. [29] Nakajima M, Eisen A, Stewart H. Comparison of corticomotoeuronal EPSPs and Macro-MUPs in amyotrophic lateral sclerosis. Muscle Nerve 1998;21:18–24. [30] Nakajima M, Eisen A, Stewart H. Diverse abnormalities of corticomotoneuronal projections in individual patients with amyotrophic lateral sclerosis. Electroencephalogr Clin Neurophysiol 1997;105:451–7. [31] Palmer E, Ashby P. Corticospinal projections to upper limb motoneurons in humans. J Physiol (Lond) 1992;448:397–411. [32] Patton HD, Amassian VE. Single- and multiple-unit analysis of cortical stage of pyramidal tract activation. J Neurophysiol 1954;17:345–63. [33] Porter R, Lemon R. Corticospinal function and voluntary movement. Monographs of the Physiological Society No. 45. Oxford: Clarendon Press, 1993: 1–421. [34] Prout AJ, Eisen A. The cortical silent period and amyotrophic lateral sclerosis. Muscle Nerve 1994;17:217–23. [35] Rothstein JD, Marleen VA, Levey AI, Martin LJ, Kuncl RW. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann Neurol 1995;38:73–84. [36] Thompson SM, Capogna M, Scanziani M. Presynaptic inhibition in the hippocampus. Trends Neurosci 1993;16:222–7. [37] Yokata T, Yoshino A, Saito Y. Double cortical stimulation in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 1996;61:596–600.