A short latency vestibulomasseteric reflex evoked by electrical stimulation over the mastoid in healthy humans

Transcription

1 J Physiol (2003), 553.1, pp DOI: /jphysiol The Physiological Society A short latency vestibulomasseteric reflex evoked by electrical stimulation over the mastoid in healthy humans Franca Deriu, Eusebio Tolu and John C. Rothwell Department of Biomedical Sciences, Section of Human Physiology and Bioengineering, University of Sassari, Sassari, Italy and Sobell Research Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, University College London, London, UK We describe EMG responses recorded in active masseter muscles following unilateral and bilateral electrical vestibular stimulation (EVS, current pulses of 5 ma intensity, 2ms duration, 3 Hz frequency). Averaged responses in unrectified masseter EMG induced by unilateral EVS were examined in 16 healthy subjects; effects induced by bilateral (transmastoid) stimulation were studied in 10 subjects. Results showed that unilateral as well as bilateral EVS induces bilaterally a clear biphasic response (onset latency ranging from 7.2 to 8.8 ms), that is of equal amplitude and latency contra- and ipsilateral to the stimulation site. In all subjects, unilateral cathodal stimulation induced a positive negative response termed p11/n15 according to its mean peak latency; the anodal stimulation induced a response of opposite polarity (n11/p15) in 11/16 subjects. Cathodal responses were significantly larger than anodal responses. Bilateral stimulation induced a p11/n15 response significantly larger than that induced by the unilateral cathodal stimulation. Recordings from single motor units showed that responses to cathodal stimulation corresponded to a brief (2 4 ms) silent period in motor unit discharge rate. The magnitude of EVS-induced masseter response was linearly related to current intensity and scaled with the mean level of EMG activity. The size of the p11/n15 response was asymmetrically modulated when subjects were tilted on both sides; in contrast head rotation did not exert any influence. Control experiments excluded a possible role of cutaneous receptors in generating the masseter response. We conclude that transmastoid electrical stimulation evokes vestibulomasseteric reflexes in healthy humans at latencies consistent with a di-trisynaptic pathway. (Received 14 May 2003; accepted after revision 28 August 2003; first published online 29 August 2003) Corresponding author F. Deriu: Department of Biomedical Sciences, Section of Human Physiology and Bioengineering, Viale San Pietro 43/b, Sassari, Italy. deriuf@uniss.it The vestibular system plays an important role in the motor control of many brainstem and postural motor systems (Lacour & Borel, 1993). Its inputs feed into eye head coordination and the head neck system by modulating the activity of eye muscles (Cohen, 1974; Precht, 1977; Berthoz, 1985), of neck muscles (Uchino et al. 1977; Bronstein, 1988; Wilson et al. 1995) and of trunk and limb extensor muscles (Pompeiano, 1972; Gernandt, 1974). Yet mechanisms whereby the vestibular inputs interact with brainstem motoneurones, other than those controlling eye muscles, are not well known. Among brainstem motoneurones, those innervating jaw-closing muscles could be a possible target of vestibular inputs, since they are muscles involved not only in chewing and speech, but also in maintaining the posture of the jaw away from its position of rest under dynamic as well as static conditions (Lund & Olsson, 1983; Miralles et al. 1987). Previous data, obtained on anaesthetised guinea-pigs, showed that a natural or an electrical stimulation, as well as a lesion of vestibular receptors affects the activity of trigeminal motoneurones innervating masseter and digastric muscles (Tolu & Pugliatti, 1993; Tolu et al. 1994, 1996; Deriu et al. 1999). They concluded that stimulation of ampullar receptors elicits polysynaptic bilateral excitatory responses in masseter and digastric motoneurones and that macular inputs exert a bilateral asymmetrical control on jaw muscles, in relation to head displacements in space. A polysynaptic pathway was suggested to be involved in connecting vestibular nuclei to trigeminal motoneurones. Up to now, very few studies have been performed on humans. Some authors (Hickenbottom et al. 1985) studied the effects of wholebody rotation on masseteric monosynaptic reflex and provided quantitative evidence that the dynamic input from vestibular ampullar receptors in response to rotation enhances masseteric motoneurone output. Other authors studied the effect of static tilt on active masseters (Deriu et al. 2000) and results suggested the existence of a bilateral and asymmetrical control exerted by macular inputs on these muscles. Several novel models of assessing vestibular reflex function have been described recently, such as click-, tap- and

2 268 F. Deriu, E. Tolu and J. C. Rothwell J Physiol electrical vestibular reflexes. These techniques activate vestibular afferents in different ways. In particular, transmastoid electrical vestibular stimulation (EVS) has been shown, in animal studies, to act upon the most distal part of the vestibular nerve, with cathodal stimulation increasing and anodal stimulation decreasing resting primary afferent discharge rate (Goldberg et al. 1984; Courjon et al. 1987; Baird et al. 1988; Nissim et al. 1994). Studies on humans support an action at the same site (Watson & Colebatch, 1997, 1998; Watson et al. 1998b). A certain degree of selectivity of low intensity electrical stimulation for otolith afferents has been reported, while higher intensity currents may additionally activate canal afferents (Zink et al. 1998). EVS has been extensively applied to the study of vestibulospinal (Britton et al. 1993; Fitzpatrick et al. 1994; Day et al. 1997; Watson & Colebatch, 1997; Séverac Cauquill & Day, 1998; Kennedy & Inglis, 2001), vestibulocollic (Watson & Colebatch, 1998; Welgampola & Colebatch, 2001; Murofuschi et al. 2002) and vestibuloocular (Pfaltz, 1967; Zink et al. 1998; Watson et al. 1998a; Kleine et al. 1999) reflexes, whereas vestibulomasseteric reflexes evoked by transmastoid electrical stimulation in humans have not been previously described. Our study was designed to investigate in healthy humans the effects induced on masseter muscle EMG activity by transmastoid EVS. Assuming such reflexes could be recorded, we wished to define their physiological properties and to determine the nature of vestibular projection to masseter motoneurones. METHODS Sixteen healthy volunteers (aged years; 6 females, 10 males), drawn from staff and fellows working in the department, were studied. The study was conducted in accordance with the ethical standards laid down in the 1964 Helsinki Declaration. The experiments were performed using techniques approved by the local ethics committee and previous informed consent was given by each subject. Experiments were carried out with the subjects seated on a comfortable chair, with the head straight, the trunk upright and the lower limbs in a semi-flexed position. EMG recording The surface EMG of both masseter muscles was recorded using standard bipolar silver/silver chloride electrodes. The simultaneous use of several configurations of the recording electrode position allowed establishment of where the largest reflex response could be detected. The largest response was generated in the lower third of the muscle, where the motor endplate region is (Godaux & Desmedt, 1975b; Widmer & Lund, 1989) and it was recorded placing the reference electrode of the pair at the level of the lower border of the mandible and the active electrode approximately 2 cm above this. The ground electrode was placed over the forehead. Both unrectified and rectified masseter EMG activities were simultaneously recorded (Digitimer D360 client Beta Version, Digitimer Ltd, Hertfordshire, England, UK), amplified (w 5000) and filtered (bandwidth Hz). EMG was sampled (5 khz) for 50 ms before to 100 ms after stimulus delivery, using a 1401 plus analog-to-digital converter (Cambridge Electronic Design LTD, Cambridge, UK) and Signal 2.10 software on a PC. The subjects were given a target level of EMG (30 % of maximal voluntary contraction, MVC) to be maintained steadily during the data collection. To this aim, a visual feedback of masseter EMG, filtered and rectified, was provided to them. In three subjects recordings from single motor units (SMU) were performed by means of a concentric needle electrode (Medelec model NDF C25, Oxford Instruments, Surrey, England, UK) into either the right (three experiments) or left (one experiment) masseter muscle, very close to where the surface electrodes were placed. The subjects were asked to make gentle contractions so that only a single or few motor units were active and they received audio-visual feedback of the firing of the SMU, enabling them to maintain a regular discharge rate. In five subjects the biting force was recorded by means of a bite bar force transducer connected to a strain gauge amplifier. The force signal was amplified and sampled for 50 ms before and 100 ms after stimulus delivery. Transmastoid electrical vestibular stimulation Transmastoid EVS was delivered by applying a direct current (constant current isolated stimulator model DS3, Digitimer Ltd, Hertfordshire, England, UK) via self-adhering 32 mm diameter electrodes (PALS, model , Nidd Valley Medical, North Yorkshire, UK) placed over the mastoid processes and secured with adhesive tape. Electrical impedance was reduced using electrode gel between the electrode and the skin. The polarity of the stimulating current was varied randomly across the trials. In the light of previous data on vestibulocollic reflexes (Watson & Colebatch, 1998; Watson et al. 1998b), electrical stimuli of 2 ms duration, 5 ma intensity and 3 Hz repetition rate were used. The use of these stimulation parameters resulted in clear EMG responses and was well tolerated by the subjects. In 16 subjects, responses to unilateral (right or left) EVS were recorded. For these experiments stimulating electrodes were placed over right or left mastoid with a second electrode on the back of the neck at the level of the seventh cervical vertebra. The stimulus polarity was changed randomly from neck positive mastoid negative (cathodal stimulation) to neck negative mastoid positive (anodal stimulation) from trial to trial. This montage allows separate stimulation of each vestibular apparatus (Watson & Colebatch, 1997). In 10 subjects, responses to bilateral EVS were recorded. A current of 5 ma intensity and 2 ms duration was passed at a frequency of 3 Hz between electrodes placed over the two mastoid processes. Both combinations of electrode placement were used, i.e. cathode left anode right and cathode right anode left. Effects of current intensity, duration and frequency In seven subjects, the current intensity of unilateral EVS was varied while the duration was kept constant at 2 ms. In each subject and for each polarity of stimulation (cathodal and anodal) the effect of current intensities ranging between 4 and 9 ma was studied. Current intensities from 2 11 ma were tested in one subject. In two subjects the current duration of unilateral EVS was varied from 1 to 3 ms while the intensity was kept constant at 5 ma. In the same subjects the effects induced by stimulus frequency variations from 1 5 Hz were tested, while intensity and duration were kept constant at 5 ma and 2 ms, respectively.

3 J Physiol Electrically evoked vestibulomasseteric reflex 269 Effects of background muscle activation In three subjects, the unilateral EVS parameters were kept constant for both stimulation polarities (2 ms, 5 ma, 3 Hz), while the level of tonic muscle activation was varied between 25 and 75 % of MVC. Frequent pauses in data collection were made for strong contractions in order to prevent fatigue. Effects of head position and of head tilt In two subjects, the responses evoked by EVS in control conditions, i.e. with the subject keeping the head straight, were compared with those recorded during the rotation of the head on both the left and right side. In five subjects, the p11/n15 response evoked by unilateral cathodal EVS at 5 ma in control conditions, i.e. with the subject keeping the trunk upright and the head straight (vertical position), was compared to that recorded during a 30 deg tilt to either the left or right side. The subjects were seated on a tilting chair with full lateral support of the head, trunk, hips and legs. After changing the tilt angle, 2 min were allowed before collection of data, in order to allow a steady state to develop and to avoid possible interference from fast-adapting receptors. We chose to deliver only cathodal stimulation since the cathodal response had been shown to be more consistent and significantly larger than the anodal response. EMG activity was recorded in the masseter ipsilateral to the side where the EVS was applied, since the ipsilateral response was usually slightly greater than the contralateral response. All five subjects were recorded with cathodal stimulation applied to the right mastoid; in two of them a second set of recordings was performed with cathodal stimulation of the left mastoid. Thus unilateral cathodal stimulation was applied in seven different experiments. Control experiments In five subjects, we investigated effects induced in active masseter EMG by electrical stimulation of various regions of the skin. These were the mastoid processes, the lateral and medial surfaces of both ear auricle and lobe, the lateral surface of the cheek just anterior to the ear tragus, and the postero-lateral surface of the neck just a little below the mastoid process. In some trials, stimulation parameters were the same as those used for mastoid electrical stimulation, in others they were adjusted to produce the same cutaneous feelings as perceived by the subjects during transmastoid electrical stimulation. Data analysis In all experiments we recorded averaged unrectified and rectified EMG responses to 800 stimuli of each type delivered at 3 Hz during both muscle contraction at the prescribed clenching level ( active trace) and during relaxation ( relaxed trace). To reduce or remove stimulus artifact, the relaxed trace was subtracted from the active one, as suggested by Watson & Colebatch (1998). The reflex responses were measured from the unrectified average, while the rectified average was used to quantify the level of tonic activation. Peaks in averaged unrectified EMG were described by using the same notation used by Watson & Colebatch (1998) to describe the EVS-induced vestibulocollic reflex, i.e. the mean latency preceded by the lower case letters p for positive peaks (downward deflection) and n for negative peaks (upward deflection). Responses that inverted in response to stimulations of opposite polarity were taken to be of vestibular origin (see Discussion). Response amplitudes in unrectified EMGs were measured peak to peak, and then divided by the mean of the rectified EMG for the 50 ms preceding the stimulus onset. This gives a value for response amplitude relative to the level of background muscle activation. Offline peristimulus time histograms (PSTH) of single unit discharges were generated using a PC attached to a laboratory interface (1401 plus, CED Electronics, Cambridge, UK) and associated in-house software. Previous studies (Colebatch & Rothwell, 1993) indicated that a large number of trials would be required to detect reliably any excitability changes and to give a stable prestimulus baseline discharge frequency. A total of presentations of each stimulus was used for each histogram, which was composed of 150 bins, each 1 ms wide, with a prestimulus period of 50 ms. The criteria for significant inhibition or excitation were those followed by Colebatch & Rothwell (1993), i.e. the presence of two adjacent bins each containing less than half or over 50 % more discharges than the prestimulus mean. All values are given as means ± S.D. Statistical comparison were made using Student s t test and linear correlation analysis. RESULTS Masseter EMG responses to unilateral and transmastoid electrical stimulation Using unilateral stimuli of 5 ma, 2 ms at 3 Hz, a clear EMG response could be seen in the mean unrectified masseter EMG in all 16 subjects. No responses could be detected in the mean rectified EMG. The responses were bilateral and their polarity depended on the stimulation polarity, in that the cathodal stimulation produced an initial positive negative response while anodal stimulation elicited an initial negative positive wave (Fig. 1A). According to their mean peak latencies, we termed the cathodal response a p11/n15 wave and the anodal response an n11/p15 wave. The responses had peak latencies and peak to peak amplitudes not significantly different in ipsilateral and contralateral muscles. Cathodal responses occurred in all subjects. They were larger than responses to anodal stimulation (P < ) which were clear in only 11 of 16 individuals (see Tables 1 and 2). Following unilateral electrical stimulation, the bite force signal showed no changes time-related to the p11/n15 response, in comparison to the prestimulus force level. In addition to p11/n15 and to n11/p15 responses, later responses were often seen, more frequently following unilateral cathodal (Fig. 1A) and bilateral (Fig. 1B) stimulations. Less frequently, these responses were also seen after unilateral anodal stimulation at the higher intensities (Fig. 2). The latency, amplitude and duration of these responses were considerably variable between subjects. They could be observed in both muscles and had similar waveforms independent of the side and polarity of stimulation. We recorded responses to bilateral stimulation (5 ma, 2 ms, 3 Hz) in 10 subjects. They were similar to the responses after unilateral cathodal stimulation, consisting of a clear consistent positive p11 wave followed by a less defined n15 wave (Fig. 1B), which often merged with the later responses described above. Although the latency of the response was not significantly different from that

4 270 F. Deriu, E. Tolu and J. C. Rothwell J Physiol induced by the unilateral cathodal stimulation (Table 1), its amplitude was significantly larger (n = 10, P < , see Table 2). Biphasic negative positive responses, similar to those observed after unilateral anodal stimulation, were never seen following transmastoid stimulation. Effects of current intensity, duration and frequency In all seven subjects, the amplitude of the masseter response (expressed as a ratio of the background EMG) to unilateral electrical stimulation increased linearly with increasing stimulation current intensity (Fig. 2). This effect was observed for both p11/n15 cathodal and n11/p15 anodal responses. Over the linear range correlation coefficients (r 2 ) were and for cathodal and anodal responses (P < in each case), respectively. In one subject studied at 11 ma, the responses were the same amplitude as at 9 ma, suggesting that the response amplitude may saturate at high stimulus intensities. However, such intensities are rather uncomfortable for most subjects and were not studied in detail. In two subjects, an increase in amplitude of the masseter response was observed when the stimulus duration increased from 1 ms to 3 ms, while the intensity was kept constant at 5 ma. In the same subjects the effect of various stimulus frequencies from 1 to 5 Hz was tested and no significant differences were found. Effects of background muscle activation In all three subjects (six sides) there was a strong linear correlation between background muscle activation and masseter response amplitude to EVS at a constant current intensity of 5 ma (Fig. 3). This relation was observed for p11/n15 cathodal as well as for n11/p15 anodal response Figure 1. Masseter EMG responses to unilateral and bilateral electrical stimulation over the mastoids A, averaged unrectified EMG responses (n = 800) to electrical stimulation (5 ma, 2 ms, 3 Hz) applied on the right mastoid process recorded from active ispi- (top traces) and contralateral (bottom traces) masseter muscles (MM), in a subject. The cathodal stimulation induced in both muscles a short-latency biphasic positive negative wave; anodal stimulation induced a response of opposite polarity. B, averaged responses (n = 800) to transmastoid electric stimulation (5 ma, 2 ms, 3 Hz) recorded from active masseter muscles, in the same subject. The responses to bilateral electrical stimulation delivered with both electrode configurations (cathode right anode left and cathode left anode right) are shown. The short-latency responses induced by unilateral cathodal and bilateral stimulation were followed by less clear later responses. Note that each trace is the result of subtracting an average obtained with MM relaxed from an average with MM activated, to minimise artifact. In all traces arrows indicate the time of stimulus onset and white gaps correspond to the duration of stimulus artifact.

5 J Physiol Electrically evoked vestibulomasseteric reflex 271 amplitudes (r 2 = 0.93 and 0.89, respectively, with P < in each case; y = 1.04x for cathodal responses and y = 0.8x _ 7.5 for anodal responses), without any evidence of saturation at the highest activation levels. Effects of head position and of head tilt No significant changes were observed in masseter responses to EVS when subjects held their head rotated to left or right sides during acquisition of the data. In contrast, tilt of the body from side to side had a significant influence on the amplitude of the p11/n15 response in all seven experiments. The amplitude of the response was measured in the masseter ipsilateral to a unilateral cathodal stimulus. Taking the mean amplitude (expressed as a ratio of the background EMG) of the response elicited with the subjects upright as 100 %, the amplitude of the p11/n15 response was 143 % (P < 0.01, n = 7) during tilt of the body towards the side of stimulation; in contrast, it was only 71 % during tilt away from the side of stimulation (P < 0.03, n = 7). Control experiments In five subjects electrical stimulation of the skin overlying: (i) the mastoid processes, (ii) the lateral and medial surfaces of both ear auricle and lobule and (iii) the postero-lateral surface of the neck, just a little below the mastoid process, elicited no responses in masseter muscles voluntarily activated at 30 % MVC. In contrast, stimulation of the skin anterior to the ear tragus was able to evoke a bilateral inhibitory response, visible in the rectified EMG, which had an onset of ms and lasted 7 10 ms. Figure 2. Relationship between the amplitude of masseter response to electrical vestibular stimulation and stimulation current intensity A, selection of averaged (n = 800) unrectified EMG traces, recorded from a single subject, showing the responses of the left masseter muscle, activated at 30 % of MVC, to ipsilateral electrical stimulation (2 ms, 3 Hz) at increasing intensity. The amplitude of both p11/n15 cathodal and n11/p15 anodal responses, increased with increasing stimulation intensity values, reported on the left. In all traces arrows indicate the time of stimulus onset and white gaps correspond to the duration of stimulus artifact. B, graph of mean ± S.D. amplitude of cathodal and anodal responses (expressed as a ratio of the background EMG) versus the stimulation intensity values, obtained from all seven subjects.

6 272 F. Deriu, E. Tolu and J. C. Rothwell J Physiol Figure 3. Relationship between the amplitude of masseter response and level of background activation A, selection of averaged (n = 800) unrectified EMG traces, recorded from a single subject, showing the responses of both masseter muscles to cathodal electrical stimulation (5 ma, 2 ms, 3 Hz) applied to the right mastoid process. The level of prestimulus mean rectified EMG for each muscle is shown on the left side of each EMG trace. In all traces arrows indicate the time of stimulus onset and white gaps correspond to the duration of stimulus artifact. B, graph of response amplitude (peak to peak amplitude, uncorrected for different levels of activation) versus mean rectified EMG. Circles and diamonds correspond to single subjects (n = 3, 6 sides) each one performing three different levels of muscle activation during cathodal and anodal stimulation, respectively. The lines are fitted linear regression lines (see Results). Figure 4. Response of a masseter single motor unit to unilateral electrical stimulation over the mastoid PSTHs of the response of the same SMU recorded from the right masseter muscle to 800 consecutive stimulations (5 ma, 2 ms, 3 Hz) applied in turn to ipsi- and contralateral mastoid processes. The effects induced by cathodal and anodal stimulation are shown. Stimuli were given at the time indicated by the arrow. Gaps filled with dots, starting at time 0 ms, correspond to the stimulus artifact. Bin duration = 1 ms.

7 J Physiol Electrically evoked vestibulomasseteric reflex 273 SMU responses to unilateral electrical stimulation Figure 4 illustrates the data from a single motor unit in the right masseter. Cathodal stimulation (5 ma, 2 ms, 3 Hz) produced a clear inhibition of unit firing lasting 2 4 ms (mean onset latency 10.8 ± 0.6 ms; mean duration 3.0 ± 0.8 ms; n = 4 units), whether the cathode was ipsilateral or contralateral to the side of the unit. There was no consistent effect after anodal stimulation. The effect induced by unilateral electrical stimulation at two different intensities was tested on one of the four SMUs (Fig. 5). Increasing the intensity from 7 to 9 ma induced an increase in duration of the inhibitory period from 3 to 5 ms; this effect was observed only following the cathodal stimulation. DISCUSSION The present results have shown that unilateral as well as bilateral electrical stimulation of mastoids evokes reproducible, short-latency, short-duration, biphasic responses in averaged unrectified EMG recorded from active masseter muscles of healthy humans. These responses were observed bilaterally following unilateral stimulation and were of opposite polarity depending on whether a cathodal or an anodal stimulation was given. Control experiments showed that these responses were not due to the stimulation of cutaneous receptors, and we argue below that they are consistent with a reflex response evoked by activation of vestibular afferent fibres. Recordings from single masseter motor units showed that cathodal stimulation produces a brief, short latency inhibition of low-threshold masseter motoneurones, whereas inconsistent effects were noted after anodal stimulation. Later responses were often seen after unilateral cathodal or bilateral stimulation and more rarely after unilateral anodal stimulation at high intensity. In the present study, we focused our attention on the short-latency masseter response, which has never been described previously. Origin of the electrically induced masseter responses Animal studies using long-lasting (several seconds in duration: galvanic ) electrical stimulation have shown that such stimuli affect the vestibular system by modulating the continuous firing level of peripheral vestibular afferents, with the firing rate of afferents being increased on the side of the cathode and decreased on the side of the anode (Goldberg et al. 1984; Courjon et al. 1987; Baird et al. 1988; Nissim et al. 1994). Watson & Colebatch (1998) have argued that galvanic stimulation with stimuli ranging in duration from 2 ms (as used in the present experiments) to several seconds has essentially the same effect in humans. In addition, it has been shown in humans that cathodal and anodal stimulation results in opposite Figure 5. Response of a masseter single motor unit to unilateral electrical stimulation at increasing current intensity values PSTHs of the response of the same SMU recorded from the right masseter muscle to 800 consecutive stimulations applied to the contralateral mastoid process. The effects induced by cathodal and anodal stimulation at intensity values of 7 ma (upper histograms) and of 9 ma (lower histograms) are shown. Stimuli were given at the time indicated by the arrows. Gaps filled with dots, starting at time 0 ms, correspond to the stimulus artifact. Bin duration = 1 ms.

8 274 F. Deriu, E. Tolu and J. C. Rothwell J Physiol effects on eye movements (Surczynsky & Ernst, 1989; Watson et al. 1998a), postural sway (Watanabe et al. 1989; Britton et al. 1993; Day et al. 1997), limb (Watson & Colebatch 1997) and neck (Watson and Colebatch 1998; Watson et al. 1998b) muscle activity. Our findings of opposite effects for cathodal and anodal stimulation are consistent with a vestibular origin of the masseter responses we have described. Furthermore, the asymmetric modulation of the p11/n15 cathodal response observed when an additional asymmetric vestibular stimulation was given by tilting the subjects, provides further support to the idea that the response was of vestibular origin. It is possible that masseter muscle spindles were affected asymmetrically by body tilt, but this is likely to be a minor factor in the present experiments since subjects were performing a steady isometric contraction with negligible or no jaw displacement. A possible contribution of cutaneous receptors, which could have been activated by the stimuli, was excluded by control experiments demonstrating no masseter responses following non-noxious and noxious electrical stimulation of the skin overlying the mastoid process, the lateral and medial surfaces of the external ear and the postero-lateral surface of the neck. Only if the cutaneous stimulation was applied on a cutaneous area supplied by trigeminal sensory fibres were we able to induce bilaterally a strong inhibitory response. However this differed from the response described above, both in beginning later (17 20 ms) and lasting longer (7 10 ms). We suggest this is a cutaneous reflex, which has been studied frequently by others (Godaux & Desmedt, 1975a; Lund et al. 1983; Romaniello et al. 2000; Deriu et al. 2003) in masseter muscles. In our study, as well as in others (Bucher et al. 1998; Bense et al. 2001) the subjects did not perceive any acoustic sensations during electrical stimulation. Indeed in human studies, transmastoid electrical stimulation has been assumed to activate vestibular rather than acoustic afferents, hence our use of the term electrical vestibular stimulation (EVS). It has been used as a means of investigating many functions and disorders of the vestibular system and to characterise vestibulardependent responses in the eye (Pfaltz, 1967; Zink et al. 1998; Watson et al. 1998a; Kleine et al. 1999), neck (Watson & Colebatch, 1998; Watson et al. 1998b; Welgampola & Colebatch, 2001; Murofuschi et al. 2002), arm and leg (Britton et al. 1993; Fitzpatrick et al. 1994; Day et al. 1997; Watson & Colebatch, 1997; Séverac Cauquill & Day, 1998; Kennedy & Inglis, 2001) muscles. However, whether electrical stimulation at the mastoid level also has an action on the acoustic system is still unclear, since it seems possible that electrical stimuli can spread from the mastoid to the inner ear and neighbouring tissues and thereby affect acoustic receptors and/or fibres in addition to the vestibular ones. To our knowledge, no animal investigations have been performed on possible effects induced by electrical stimulation at the mastoid level on either cochlear receptors, or acoustic fibres travelling through the eighth nerve or cochlear nuclei. Data from cerebral functional magnetic resonance imaging (fmri) studies in humans, carried out during the application of currents to the mastoids, have given partial and contradictory answers. Bucher et al. (1998) performed a fast low-angle shot (FLASH) study which showed significant activations of the parieto-insular-vestibular cortex (PIVC) and the posterior medial thalamus, which according to monkey studies are involved in processing vestibular function (Guldin & Grüsser, 1998). However, simultaneous activation of the transverse temporal (Heschl s) gyrus indicated an involvement of the auditory cortex. At about the same time, Lobel et al. (1998) carried out a whole-brain study and reported that most activation foci could be related to the vestibular component of the stimulus, being located in cerebral areas equivalent to the monkey s PIVC region (Grüsser et al. 1983, 1990; Guldin & Grüsser, 1998). In particular, Lobel et al. (1998) highlighted the fact that activation foci were located in an area clearly distinct from the auditory cortex. Finally, a more recent whole-brain study (Bense et al. 2001), investigated in detail the differential activation patterns induced by mastoid electrical stimulation and by purely auditory stimulation, and the results supported the view that EVS only represents a weak stimulus for the auditory system. This is not to deny the existence of sound-induced reflexes in human jaw muscles. Indeed, the response of masseter muscles to auditory stimuli (high-intensity clicks or tone bursts) has been described by Meier-Ewert et al. (1974) as an acoustic jaw-reflex, which consists of a silent period in the active masseter EMG. This reflex is very similar to the inion response first described by Bickford et al. (1964). However, it differs from the EVS-induced masseter responses we have described in this study both in beginning later (14 ms) and lasting longer (11 ms). In the light of the above reported data, we conclude that it is very unlikely that the short-latency, short-duration responses we have described following mastoid electrical stimulation are cochlear in origin; rather they may represent a vestibulomasseteric reflex. It is interesting to note that anodal responses were smaller and less consistent than cathodal responses. In addition, responses to bilateral (transmastoid) stimulation, which are likely to be the summation of simultaneous actions at the cathode and the anode, were always the same polarity as for unilateral cathodal rather than unilateral anodal stimulation. These data suggest that cathodal stimulation is more effective than anodal stimulation in eliciting the vestibulomasseteric reflex. In some previous reports on unilateral mastoid electrical stimulation in humans, no

9 J Physiol Electrically evoked vestibulomasseteric reflex 275 explicit comparison between the magnitudes of cathodal and anodal effects has been given (Surczynsky & Ernst, 1989; Watanabe et al. 1989; Cass et al. 1996); in others cathodal and anodal effects of similar amplitude were described (Weiss & Tole, 1973; Furman & Cass, 1996). Other authors reported larger effects induced by cathodal stimulation in comparison to those induced by anodal stimulation on leg (Watson & Colebatch, 1997) and neck (Watson & Colebatch, 1998) muscles. We suggest that the larger effect of cathodal stimulation in the present experiments is due to two factors. The first relates to the differences in action of anodal and cathodal stimulation on the vestibular nerve. Goldberg et al. (1984) found that anodal stimulation could completely suppress the ongoing discharge of vestibular afferents, effectively limiting the magnitude of its effect. In contrast they found no clear evidence for response saturation in the excitatory response to cathodal stimulation. The implication is that cathodal stimulation may be able to produce a greater change in discharge rate than anodal stimulation. In addition to the greater possible magnitude of change, cathodal stimulation, by activating afferent axons, will also produce a more abrupt change in discharge than anodal stimulation, which can only silence ongoing activity. Given that the responses we observed have such a short latency and short duration, the combination of a larger, more abrupt change in afferent firing may make cathodal stimulation more effective than anodal stimulation. Nature of the electrically induced masseter responses The reflex responses we have described have rather unusual characteristics: (i) they are visible as a biphasic wave in average unrectified EMGs but (ii) are not seen in average rectified EMGs. The transmastoid electrically induced vestibulocollic reflex (Watson & Colebatch, 1998) and the trigeminocollic reflex (Di Lazzaro et al. 1995) have similar characteristics. The reason for this is as follows. As pointed out by Colebatch & Rothwell (1993), the average of a series of randomly selected unrectified EMG traces is flat because there is, on average, cancellation of the positive and negative phases of motor unit action potentials. That is, the negative part of an action potential in one sweep will be cancelled by the positive part of an action potential in another sweep, since the action potentials occur by chance at different times in each sweep. In the present experiments, the single unit data show that cathodal stimulation produces an abrupt inhibition of unit firing that lasts only 2 3 ms. Importantly, this occurs at the same time in every sweep. Thus while action potentials occur at random times over most of each sweep, no action potentials ever begin over these 2 ms in every sweep. The fact that these action potentials are missing means that there is no cancellation of the positive tails of action potentials that start before the 2 ms period of inhibition. Similarly, there is nothing to cancel the negative peaks of action potentials that start just after the 2 ms period of inhibition. There is thus a failure of cancellation because action potentials are always missing at the same time in every sweep. Effectively what is seen in the average unrectified EMG is a missing action potential that has the opposite polarity to the normal motor unit potential. The reason that this does not appear in the rectified trace is because the period of inhibition is so much shorter than the duration of a typical action potential. We propose that the p11/n15 response to cathodal stimulation is produced by synaptic activation of an inhibitory projection to masseter motoneurones by a sudden discharge in the vestibular nerve. Given the short onset latency ( ms) and the abrupt onset of inhibition, we believe this must involve only two or three synaptic relays, as illustrated in Fig. 6. The vestibulotrigeminal connection could be mediated by an inhibitory interneurone localised somewhere in the brainstem. Animal experiments showed that relay stations involved in reciprocal anatomical connections between vestibular and trigeminal nuclei are the pontomedullary reticular formation, the trigeminal sensory complex, the cervical spinal cord and premotoneurones localised in the supratrigeminal area (Kolta, 1997; Buisseret-Delmas et al. 1999; Xiong & Matsushita, 2000; Valla et al. 2003). The response to anodal stimulation, which decreases ongoing vestibular nerve discharge, has the opposite effect to cathodal stimulation: it reduces any ongoing tonic inhibition and excites masseter motoneurones. Thus, the EMG response is negative positive, like the muscle M-wave. Reducing inhibition for a short period is likely to have a smaller effect on ongoing EMG activity than the abrupt onset of extra inhibition (as produced by cathodal stimulation), and may explain why anodal effects were not as clear as cathodal effects. Indeed, we did not see any evidence for excitation in our single motor unit recordings after anodal stimulation. This may have been because our sample was small and anodal responses were not clear in all subjects, or because anodal stimulation preferentially recruits high-threshold motor units that we could not readily sample with conventional needle EMG. Vestibular influences on trigeminal motor system Most of the studies on the role played by the vestibular system on motor control mechanisms have been focused on vestibular-dependent responses exhibited by eye, neck, trunk and limb muscles aimed at the maintenance of body posture and balance (Pompeiano, 1972; Gernandt, 1974; Precht, 1977; Uchino et al. 1977; Cohen, 1974; Berthoz, 1985; Bronstein, 1988; Britton et al. 1993; Lacour & Borel, 1993; Fitzpatrick et al. 1994; Wilson et al. 1995; Day et al. 1997; Watson & Colebatch, 1997; Séverac Cauquill & Day, 1998; Watson & Colebatch, 1998; Watson et al. 1998b). In contrast, very few data are available in the literature on vestibulo-trigeminal relationships; nevertheless they provide clear evidence that the vestibular system can

10 276 F. Deriu, E. Tolu and J. C. Rothwell J Physiol modulate the activity of trigeminal motoneurones. Experiments on anaesthetised guinea-pigs have shown that, in basal conditions, masseter muscles are submitted to a tonic excitatory control from the vestibular system in the same way as in other antigravity muscles. Although bilaterally expressed, this control has been reported to be asymmetric, being stronger on the contralateral muscle (Tolu & Pugliatti, 1993; Tolu et al. 1994). The same authors showed that the electrical stimulation of ampullar receptors induced bilaterally in single masseter and digastric motoneurones (Tolu et al. 1996; Deriu et al. 1999) excitatory long-lasting (14 16 ms) responses at latencies (8 12 ms) consistent with a polysynaptic transmission. To our knowledge, only two studies have been performed on humans and they confirmed the vestibulo-trigeminal relationship evidenced in the animal model. Hickenbottom et al. (1985) studied the effects of whole-body rotation on the masseteric monosynaptic reflex, and results provided evidence that the dynamic input from vestibular ampullar receptors in response to rotation enhances masseteric motoneurone output. Other authors (Deriu et al. 2000) showed that static tilt induces bilateral asymmetric changes in masseter muscle EMG activity, consisting of an increase in the muscle Figure 6. Proposed reflex pathway subserving masseter short-latency responses to electrical vestibular stimulation The masseter response to EVS could be generated on the ipsilateral masseter muscle by an uncrossed pathway; the same response is generated on the contralateral muscle by a crossed pathway. This vestibulotrigeminal connection could be mediated by an inhibitory interneurone located in a relay station at the brainstem level. Possible brainstem relay stations are indicated in the figure.

11 J Physiol Electrically evoked vestibulomasseteric reflex 277 contralateral to the side of the tilt and a decrease in the ispilateral muscle. At first sight, the present results appear to contradict these previous data. Our reflex is bilateral, symmetrical and predominantly inhibitory rather than being asymmetric and excitatory. In addition it has a short latency and duration, rather than being prolonged and late. However, such a situation is not unique. For example, mastoid EVS evokes two sets of responses in leg and arm muscles (Britton et al. 1993; Fitzpatrick et al. 1994; Watson & Colebatch, 1997): a short-latency, short-duration response followed by a longer lasting response of opposite polarity. The latter sustained effect is more powerful and is responsible for the postural sway that mastoid EVS evokes in standing subjects. We propose that a similar organisation may occur in the vestibular connections to masseters: indirect pathways that operate in a tonic fashion on which previous studies have been focused and a short-latency pathway, preferentially activated by phasic inputs, which the present experiments have explored. Bite force recordings failed to show any change in force signal during the early short period in which the p11/n15 cathodal response occurred. This result was not surprising considering the small duration and small size of the EMG response. Some changes in force might be expected to occur when later and long-lasting EMG responses to EVS occur. However, later responses were not investigated in the present study and force recordings were not made in previous studies focused on long-latency, long-lasting trigeminal responses to vestibular stimulation (Tolu & Pugliatti, 1993; Tolu et al. 1996). Short-latency masseter responses we have described are similar in onset and peak latencies, as well in their inhibitory nature, to the EVS-induced vestibulocollic reflex described by Watson & Colebatch (1998) in sternomastoid muscles (SCM). Like the vestibulocollic reflex, the magnitude of the masseter response was linearly related to current intensity and scaled with the mean level of EMG activity. In contrast, several differences related to the stimulation recording side exist. Cathode-induced masseter responses were bilateral and of the same polarity (positive negative) on each side, while cathodal vestibulocollic responses were positive negative in the ipsilateral SCM and negative positive in the contralateral SCM. Furthermore, anodal stimulation produced a negative positive wave on both masseter muscles, whereas it induced a response of the same polarity only in the SCM ipsilateral to the anode (Watson & Colebatch, 1998). These differences are probably related to the different functional role played by these muscles, SCMs working as antagonist muscles in head rotation, masseters operating together on the mandible to position it appropriately via its sliding-hinge joint. In conclusion, we have described a new short-latency reflex in the human masseter muscles produced by electrical stimulation over the mastoid processes. We argue that this is a vestibulomasseteric reflex in a di- or trisynaptic brainstem pathway. This might allow vestibular inputs rapid access to jaw muscle control and allow fine tuning of voluntary motor output to masseter muscles. REFERENCES Baird RA, Desmadryl G, Fernandez C & Goldberg JM (1988). The vestibular nerve of the chinchilla. II. Relation between afferent response properties and peripheral innervation patterns in the semicircular canals. J Neurophysiol 60, Bense S, Stephan T, Yousry TA, Brandt T & Dieterich M (2001). Multisensory cortical signal increases and decreases during vestibular galvanic stimulation (fmri). J Neurophysiol 85, Berthoz A. (1985). Adaptive mechanisms in eye-head coordination. In Rev Oculomotor Control. Adaptative Mechanism in Gaze Control. Facts and Theories, ed Berthoz A & Melvill Jones G, vol. 1, pp Elsevier, Amsterdam. Bickford RG, Jacobson JL & Cody D TH (1964). Nature of average evoked potentials to sound and other stimuli in man. Ann N Y Acad Sci 112, Britton TC, Day BL, Brown P, Rothwell JC, Thompson PD & Marsden CD (1993). Postural electromyographic responses in the arm and leg following galvanic vestibular stimulation in man. Exp Brain Res 94, Bronstein AM (1988). Evidence of vestibular input contributing to dynamic head stabilization in man. Acta Otolaryngol 105, 1 6. Bucher SF, Dieterich M, Wiesmann M, Weiss A, Zink R, Yousry TA & Brandt T (1998). Cerebral functional magnetic resonance imaging of vestibular, auditory, and nociceptive areas during galvanic stimulation. Ann Neurol 44, Buisseret-Delmas C, Compoint C, Delfini C, & Buisseret P. (1999). Organisation of reciprocal connections between trigeminal and vestibular nuclei in the rat. J Comp Neurol 409, Cass SP, Redfern MS, Furman JM & Dipasquale JJ (1996). Galvanicinduced postural movements as a test of vestibular functions in humans. Laryngoscope 106, Cohen B (1974). The vestibulo-ocular reflex arc. In Handbook of Sensory Physiology, ed. Kornhuber HH, vol. 6, part 1, pp , Springer-Verlag, New York. Colebatch JG & Rothwell JC (1993). Vestibular-evoked EMG responses in human neck muscles. J Physiol 473, 18P. Courjon JH, Precht W & Sirkin DW (1987). Vestibular nerve and nuclei unit responses and eye movement responses to repetitive galvanic stimulation of the labyrinth in the rat. Exp Brain Res 66, Day BL, Séverac Cauquil A, Bartolomei L, Pastor MA & Lyon IN (1997). Human body-segment tilts induced by galvanic stimulation: a vestibularly driven balance protection mechanism. J Physiol 500, Deriu F, Milia M, Podda MV, Giaconi E & Tolu E (2003). Modulation of masseter exteroceptive suppression by nonnociceptive upper limb afferent activation in humans. Exp Brain Res 150, Deriu F, Podda MV, Chessa G & Tolu E (1999). Trigeminal integration of vestibular and forelimb nerve inputs. Arch Ital Biol 137,

12 278 F. Deriu, E. Tolu and J. C. Rothwell J Physiol Deriu F, Podda MV, Milia M, Chessa G, Sau G, Pastorino M, Aiello I & Tolu E (2000). Masseter muscle activity during vestibular stimulation in man. Arch Ital Biol 138, Di Lazzaro V, Quartarone A, Higuchi K & Rothwell JC (1995). Shortlatency trigemino-cervical reflexes in man. Exp Brain Res 102, Fitzpatrick R, Burke D & Gandevia SC (1994). Task-dependent reflex responses and movement illusions evoked by galvanic vestibular stimulation in standing humans. J Physiol 478, Furman JM & Cass SP (1996). Laboratory evaluation. I. Electronystagmography and rotational testing. In Disorders of the Vestibular System, ed. Baloh RW & Halmagyi GM. Oxford University Press, New York. Gernandt BE (1974). Vestibulo-spinal mechanisms. In Handbook of Sensory Physiology, ed. Kornhuber HH, vol. 6, part 1, pp , Springer-Verlag, New York. Godaux E & Desmedt JE (1975a). Exteroceptive suppression and motor control of the masseter and temporalis muscles in normal man. Brain Res 85, Godaux E & Desmedt JE (1975b). Human masseter muscle: H- and tendon reflexes. Their paradoxical potentiation by muscle vibration. Arch Neurol 32, Goldberg JM, Smith CE & Fernandez C (1984). Relation between discharge regularity and responses to externally applied galvanic currents in vestibular nerve afferents of the squirrel monkey. J Neurophysiol 51, Grüsser O-J, Pause M & Schreiter U (1983). A new vestibular area in the primate cortex. Soc Neurosci Abstr 9, 749. Grüsser O-J, Pause M & Schreiter U (1990). Localization and responses of neurons in the parieto-insular vestibular cortex of awake monkeys (Macaca fascicularis). J Physiol 430, Guldin WO & Grüsser O-J (1998). Is there a vestibular cortex? Trends Neurosci 21, Hickenbottom RS, Bishop B & Moriarty MT (1985). Effects of whole body rotation on masseteric motoneuron excitability. Exp Neurol 89, Kennedy PM & Inglis JT (2001). Modulation of the soleus H-reflex in prone human subjects using galvanic vestibular stimulation. Clin Neurophysiol 112, Kleine JF, Guldin WO & Clarke AH (1999). Variable otolith contribution to the galvanically induced vestibulo-ocular reflex. Neuroreport 10, Kolta A (1997). In vitro investigation of synaptic relations between interneurons surrounding the trigeminal motor nucleus and masseteric motoneurons. J Neurophysiol 78, Lacour M & Borel L (1993). Vestibular control of posture and gate. Arch Ital Biol 131, Lobel E, Kleine JF, Le Bihan D, Leroy-Willig A & Berthoz A (1998). Functional MRI of galvanic vestibular stimulation. J Neurophysiol 80, Lund JP, Lamarre Y, Lavigne G & Duquet G (1983). Human jaw reflexes. In Motor Control Mechanisms in Health and Disease, ed. Desmedt JE, pp , Raven Press, New York. Lund JP & Olsson KA (1983). The importance of reflexes and their control during jaw movement. Trends Neurosci 11, Meier-Ewert K, Gleitsmann K & Reiter F (1974). Acoustic jaw reflex in man: its relationship to other brainstem and microreflexes. Electroencephalogr Clin Neurophysiol 36, Miralles R, Manns A, Nass X, Pasini C & Rocabado M (1987). Influence of protrusive functions on electromyographic activity of elevator muscles. Cranio 5, Murofuschi T, Takegoshi H, Ohki M & Ozeki H (2002). Galvanicevoked myogenic responses in patients with an absence of clickevoked vestibulocollic reflexes. Clin Neurophysol 113, Nissim KR, Cass SP & Scudder CA (1994). Galvanic sensitivity of monkey vestibular afferents stimulated with transmastoid currents. Otolaryngol Head Neck Surg 111, 150. Pfaltz CR (1967). Nystagmographyc study of the galvanic vestibular response. Electroencephalogr Clin Neurophysiol 23, 584. Pompeiano O (1972). Vestibulo-spinal relations: vestibular influences on gamma motoneurons and primary afferents. In Progress in Brain Research. Basic Aspect of Central Vestibular Mechanisms, ed. Brodal A & Pompeiano O, vol. 37, pp , Elsevier, New York. Precht, W (1977). The functional synaptology of brainstem oculomotor pathways. In Control of Gaze by Brain Stem Neurons, ed. Baker R & Berthoz A, pp , Biomedical Press, Amsterdam, New York and Elsevier, North-Holland. Romaniello A, Svensson P, Cruccu G & Arendt-Nielsen L (2000). Modulation of exteroceptive suppression periods in human jawclosing muscles induced by summation of nociceptive and nonnociceptive inputs. Exp Brain Res 132, Séverac Cauquil A & Day BL (1998). Galvanic vestibular stimulation modulates voluntary movement of the human upper body. J Physiol 513, Surczynsky W & Ernst A (1989). Galvanic testing in neuro-otological diagnosis. Acta Otolaryngol 468, Tolu E, Caria MA, Chessa G, Melis F, Simula ME, Podda MV, Solinas A & Deriu F (1996). Trigeminal motoneurone responses to vestibular stimulation in the guinea pig. Arch Ital Biol 134, Tolu E & Pugliatti M (1993). The vestibular system modulates masseter muscle activity. J Vest Res 3, Tolu E, Pugliatti M, Lacana P, Chessa G, Caria MA & Simula ME (1994). Vestibular and somatosensory afferents modulate masseter muscle activity. J Vest Res 4, Uchino Y, Sato H, Sasaki M, Imagawa M, Ikegami H, Isu N & Graf W (1997). Sacculocollic reflex arcs in cats. J Neurophysiol 77, Valla J, Delfini C, Diagne M, Pinganaud G, Buisseret P & Buisseret- Delmas C (2003). Vestibulotrigeminal and vestibulospinal projections in rats: retrograde tracing coupled to glutamic acid decarboxylase immunoreactivity. Neurosci Lett 340, Watanabe Y, Mizukoshi K, Ohi H, Yasuda K, Ohasi N & Kobayashi H (1989). Retro-labyrintine disorders detected by galvanic body sway responses in routine equilibrium examinations. Acta Otolaryngol 468, Watson SRD, Brizuela AE, Curthoys IS, Colebatch JG, Macdougall HG and Halmagyi GM (1998a). Maintained ocular torsion produced by bilateral and unilateral galvanic (DC) vestibular stimulation in humans. Exp Brain Res 122,