- Email: [email protected]
The reflex responses of single motor units in human lower lip muscles to mechanical stimulation
Brain Research, 251 (1982) 65-75 Elsevier Biomedical Press
65
The Reflex Responses of Single Motor Units in Human Lower Lip Muscles to Mechanical Stimulation MICHAEL D. McCLEAN and ANNE SMITH Department of Speech and Hearing Sciences and ( A.S.) Department of Physiology and Biophysics, University of Washington, Seattle, WA 98195 (U.S.A.)
(Accepted April 22nd, 1982) Key words: reflexes - - single motor units - - human lower lip
Single motor unit activity was recorded from human lower lip muscles while subjects maintained a firing rate of 20 ips using audio and visual feedback. Each motor unit was classified as belonging to one of three muscles, orbicularis oris inferior (OOI), mentalis (MENT), or depressor labii inferior (DLI). Reflex responses were elicited in single motor units by applying small mechanical displacements (0.5 mm) to the corner of the mouth. Reflex responses were analyzed by constructing peristimulus histograms. Of 130 units analyzed, 123 showed statistically significant changes in their probability of firing in the 50 ms period following the stimulus. Reflex responses were typically polymodal with both excitatory and suppression components present. The excitatory responses occurred at latencies consistent with previous descriptions of the perioral reflexes and were labelled El (X = 15.9 ms) and E2 (X = 35.6 ms). Reflex suppression responses (S) were also consistently observed. There was a tendency for the pattern and/or relative magnitudes of reflex response components to depend on the muscle classification of the motor unit; E1 responses predominating in OOI units, E2 responses in MENT units, and S responses in DLI units. This finding provides preliminary evidence for muscle-specificprojections between the trigeminal sensory system and facial nucleus motoneurones in humans. INTRODUCTION Electrical stimulation of trigeminal nerve branches or mechanical stimulation of the perioral tissues produces diffuse excitatory reflex responses in human lip musclesS,11, 20. Little is known, however, about the specific organization of trigeminal inputs to facial nucleus motoneurons, and there is no clear evidence that the reflex effects of mechanical stimulation are different for different perioral muscles. In order to study the perioral reflexes with quantitative mechanical stimuli and to determine whether or not there are differing patterns of effects on motoneurons innervating closely adjacent but functionally distinct muscles, we have adopted the methods used by Stephens and his colleagues4,12, zT. In this approach, subjects use visual and/or auditory feedback to maintain a constant firing rate in a motor unit, and the effects of repeated stimulation are determined by constructing a poststimulus time histogram. Changes in the probability of firing timelocked to the stimulus reflect the contribution made 0006-8993/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press
by the stimulus to the total firing pattern of the cell. Stephens and colleagues4,12, 27 have shown that innocuous mechanical stimuli produce reflex effects on motoneurons of human hand muscles. These responses were large and elicited only by skin stimulation in areas whose natural stimulation could be expected to be associated with movements involving activity in a particular muscle. Garnett and Stephens (ref. 12) suggested, therefore, that cutaneous input may play a major role in modifying motor unit discharge during movement. The analysis of reflex effects on single motor unit activity is important to the study of lip muscles because the fibers of different lip muscles are often interlaced. For example, the fibers of antagonist muscles, mentalis (a lower lip elevator) and depressor labii inferior (a lower lip depressor), blend at their common insertion on the skin of the chin in. Thus, the activity of one muscle cannot be determined with certainty from interference-pattern E M G recordings. Smith et al. an reported that even a fine-wire single motor unit electrode with a distance
66 between leads of less than 25 # m typically recorded both lower lip elevator and depressor units at a single intramuscular site in the chin. Previous descriptions of the perioral reflexes and the fact that the lip muscles are not spatially separate suggest that trigeminal inputs to the facial nucleus may not be organized with respect to individual muscles. However, on the basis of studies of human and cat spinal muscle systems1,5,12,14,15 and data on trigeminofacial connections in the rat 10, it seems possible that a muscle-specific organization does exist. The present experiment was undertaken to investigate the reflex effects of mechanical stimulation on single motor units in human lip muscles having different voluntary functions and different locations relative to the point of stimulation. The discovery of any systematic relationships between lower lip motor unit function, location, and pattern of reflex response should provide useful information on the organization of trigeminal input to the facial nucleus in humans as well as a more refined basis for experiments designed to investigate the role of the perioral reflexes in the control of oral-facial movements.
Subjects
muscle that surrounds the lower lip and functions to elevate and round the lower lip. 2. Mentalis (MENT), a muscle that arises from the mandible, inserts in the skin of the chin and functions to elevate the lower lip. 3. Depressor labii inferior (DL1), a muscle that arises from the mandible, inserts in the skin of the chin and the skin and mucosa of the lower lip, and functions to lower the lower lip. The location of each motor unit (the estimated point at which the needle ended after the insertion) was placed within a Cartesian coordinate system in which the right corner of the mouth was taken as the origin. Vertical and horizontal measurements were converted to percentage values based on the facial dimensions of the subject. This procedure allowed motor units from different subjects to be compared in terms of their relative locations. Each motor unit was classified as either a lower lip elevator or depressor on the basis of (1) its activation characteristics during isotonic speech and nonspeech movements of the lower lip as observed on a storage oscilloscope, and (2) the posturing of the lips during isometric activation associated with motor unit training. In a very few cases motor units did not clearly behave like a lower lip elevator or depressor on both of these types of gestures. These units were excluded from further analysis.
Subjects were seven adult volunteers. All were apparently healthy; one was a mild stutterer.
Muscle class(fication of motor units
MATERIALS AND METHODS
Single motor unit recording Single motor units were recorded with bipolar hooked-wire electrodes. The electrodes were made of two Teflon-insulated stainless steel wires (25 # m coated diameter) twisted tightly together. The ends of the wires were burned and the exposed tips were then trimmed back until the impedance of each wire was 30-300 kOhms measured at I kHz in saline with a Wheatstone bridge. The distance between the two leads was less than 25 #m. The electrodes were inserted into the muscle with a 30-gauge hypodermic needle which was then withdrawn. Motor unit potentials were amplified with a high impedance differential amplifier (Honeywell, Model 135). Art attempt was made to record single motor unit potentials from 3 lip muscles. 1. Orbicularis oris inferior (OOI), a sphincter-like
The close proximity and complex anatomy of the human perioral muscles make it difficult to classify single motor unit action potentials with particular muscles on the basis of electrode location. For example, DLI is known to blend with fibers from OOI and M E N T at the regions of its insertion into the lower lip and chin. Consistent with this observation is the earlier mentioned finding that both lower lip elevator and depressor motor units have been recorded over the same fine-wire electrode configuration in the chin z6. At our present stage of understanding of the perioral motor system, it seems most parsimonious to assume that such motor units represent distinct muscles rather than motor unit 'subpopulations' of one muscle. For the above reasons, strictly functional criteria were used in the present study to distinguish DLI from OOI and M E N T motor units. All motor units
67 which ftmctioned to depress the lower lip were classified as DLI units, regardless of electrode location. All motor units which functioned to elevate the lower lip were classified as either OOI or M E N T units. The OOI and M E N T motor units were distinguished from one another solely on the basis of anatomical criteria, since these muscles have a shared function (lower lip elevation) but relatively distinct anatomical locations. Classification of individual OOI and M E N T motor units was made by comparing the electrode location with previous descriptions of the anatomy of these muscles.
Mechanical stimulation Small mechanical stretches were applied to the right corner of the subject's mouth with a positioncontrolled electromagnetic servo systemlS, 20. A force transducer with a rubber-tipped cylindrical extension (diameter = 5 mm) was fixed to the end of the displacement shaft. The rubber-tipped cylinder was placed in contact with the skin at the corner of the mouth at a standing force of 20 g. The standard stimulus was a tap with a duration of 10 ms, a rise time of 4 ms, and a displacement of 0.5 mm. Such stimuli were applied at a rate of between 3 and 4 taps/s. A pre-stimulus trigger signal occurring 50 ms before the tap, a voltage analog of the shaft position, and the E M G signal were recorded on a Honeywell FM tape recorder (Model 5600). The frequency response in all 3 channels was flat from DC to 2.5 kHz.
Experimental procedure Subjects were seated in a dental chair with their heads supported by a comfortable headrest. The single motor unit signal was fed back by means of a loudspeaker and an oscilloscope to aid subjects in controlling motor unit firing rate. The oscilloscope was triggered on the motor unit potential and the time base set so that subjects could maintain a rate of 20 ips. After the subject had practiced long enough to maintain relatively steady rates of discharge, stimuli were delivered to the corner of the mouth. The force signal was displayed on one channel of the oscilloscope to aid in the maintenance of the standing force which varied with 'drift' in head position. Motor unit activation and stimulation were sustained for a period of at least 3 min in
order that a minimum of 512 stimuli could be applied.
Data analysis Peristimulus time histograms were constructed for each motor unit with an Ortcc Time Histogram Analyzer (Model 4620A-4621). The bin width was 1 ms, and a 255 ms period was analyzed including a 50 ms pre-stimulus interval. During the generation of a histogram each motor unit was monitored on an oscilloscope to insure that only one unit triggered the computer. In some cases, when more than one unit was active, a Time-Amplitude Window Discriminator (Bak Electronics, Model DIS-1) was used to isolate smaller amplitude motor unit potentials 2. The contents of the Ortec memory were read into a Tektronix 4051 computer for signal processing, measurement, and statistical analysis. Reflex responses were initially identified by inspecting the peristimulus histograms on the computer terminal screen. A cumulative sum plot 9 was displayed below each histogram, and it was used to make a more precise determination of the temporal onset and offset of individual responses. An interactive computer program was used to impose time markers at the points of visually detectable continuous changes in the slope of the cumulative sum plot. The computer then automatically calculated the latency, duration, magnitude and statistical significance of each response. The magnitude of reflex responses was quantified as additional pulses per stimulus. This number was calculated by subtracting the prestimulus control level from the total number of spikes occurring over the duration of the response and dividing by the number of stimuli. The statistical significance of individual responses was assessed with Poisson statistics as described elsewhere 12. Only responses reaching significance at the 5 ~ level in a two-tailed test are reported in the present study. RESULTS Histograms were constructed on a total of 130 single motor units: 38 OOI units, 43 M E N T units and 49 DLI units. Fig. 1 shows the estimated locations of the motor units. Of the 130 units, 123 showed a significant change in probability of firing in the 50 ms period following the onset of the
68
O
MENT
•
DLI
8
001
Fig. 1. An illustration of the estimated locations of 122 motor units analyzed in the present study. stimulus. Significant responses sometimes occurred at poststimttlus latencies longer than 50 ms; however, these longer latency responses are not considered in the present report. Fig. 2 illustrates for an OO1 unit the graphic display used to measure the temporal onset and offset of the component responses for each motor unit, The top trace is a peristimulus time histogram, and the bottom trace is the cumulative sum plot calculated on this histogram. The mean firing rate of the motor units studied, as estimated from the bin counts in the 50 ms prestimulus interval, was 18.4 impulses/s (S.D. = 4.4).
Classification of motor unit responses A wide variety of polymodal response patterns was observed in the histograms of the various motor units studied. In order to describe these effects, a system was developed for assigning each motor unit to a unique class based on the temporal characteristics of the excitatory artd suppression phases in its peristirnulus histogram. The reflex responses of the
lip muscles have been described as having two excitatory components with latertcies of 10-17 ms and 25-45 ms s,11,2°. The poststimulus latencies of the excitatory responses observed in the present study were consistent with this description. These latencies were bimodally distributed with a first component response (El) having a range of 8-21 ms ('~ = 15.9, S.D. ~- 3.8, n ~ 51) and a second component response (E2) having a range of 21-50 ms ( ~ = 35.6, S.D. = 7.2, n ---- 105). The latencies of the suppression responses (S) obtained in the present study were unimodally distributed with a range of 13-50 ms (.~ ---- 28.2, S.D. -~ 8.3, n = 98). On the basis of these latency distributions, the excitatory and suppression responses in each histogram were classified as El, E2 or S responses. Eighteen motor unit response classes were found; however, 10 of these classes contained only one or two motor units and together accounted for less than 10 ~ of all the units. In order to simplify the classification system, response classes containing only one or two motor
69 1BB
f PSTH
C
N 50 T
S
t
s
E P R
I
D HE
EMcS 201 -20e .
'
"
.
.
.
"
,
e
.
.
•
.
,
.
.
~e
.
.
.
.
.
.
,
TIME(MSEC) .
.
.
.
tee
.
.
,
.
.
.
.
ISe
.
.
,
.
•
.
.
.
.
2~e
Fig. 2. A peristimulus time histogram (PSTH) and cumulativesum plot (CUMSUM) on an OOI motor unit activated at an average firing rate of 26 impulses/s. The 3 types of response components identified in the present study are indicated by the labels El, E2 and S. See the text for further explanation of this classification. units were combined with other response classes. This involved, for example, combining a category such as E1-S-E2-S with El-S-E2. Table I shows the classification of response pat-
terns of the motor units recorded from the 3 muscles. The number in each cell is the percentage of motor units within a given muscle class that showed the associated response pattern.
TABLE I
R e f l e x responses in lip elevator m o t o r units
Percentages of 001, M E N T and DLI motor units showing the various histogram response patterns
See text for explanation of the response-pattern categories. NR indicates that no significant responses occurred in the 50 ms poststimulus period.
El-E2 EI-S El-S-E2 E2 E2-S S-E2 S NR
00I (n = 38)
MENT (n = 43)
DLI (n = 49)
39 26 16 13 0 0 3 3
9 7 9 21 7 33 2 12
6 8 0 6 24 41 12 2
Table I indicates that differences exist in the reflex response patterns of the two groups of elevator motor units, OOI and MENT. The percentage of OOI units showing E1 response patterns was greater than the percentage of M E N T units with E1 response patterns. The M E N T units, on the other hand, showed a greater percentage of E2, S-E2, and E2-S patterns compared to the OOI units. This difference in the response patterns of OOI and M E N T motor units is further illustrated in Fig. 3 which shows peristimulus histograms obtained on 5 OOI units (left column) and 5 M E N T units (right column) in the same subject. It may also be seen in Fig. 3 that the OOI units had larger El responses compared with the E2
70
Fig. 3. Peristimulus time histograms obtained on 5 0 0 I motor units (on the left) and 5 MENT units (on the righ0 in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus. responses, and the M E N T units had larger E2 responses compared with the E1 responses. This difference represents a general trend seen for all the O O I and M E N T units, and it is further illustrated in Table II. Table II shows the percentage of m o t o r units in each muscle class in which the largest reflex response was El, E2 or S. TABLE II The percentage o f motor units in each muscle class in which the largest reflex response was El, E2, or S N R indicates that no significant 50 ms poststimulus time interval.
E1 E2 S NR
responses occurred in the
001 (n : 38)
MENT (n = 43)
DLI (n = 49)
63 29 5 3
9 72 7 12
2 51 45 2
R e f l e x r e s p o n s e s in lip d e p r e s s o r m o t o r units
Table I indicates that the D L I motor units had a greater percentage of E2-S, S-E2, and S response patterns compared to the elevator units (OOI and MENT). As with the M E N T units, the D L I units showed few E1 response patterns relative to the OOI units. The differences in the reflex responses of the depressor and elevator motor units are further illustrated in Figs. 4 and 5. Fig. 4 shows the peristimulus histograms of 5 OOI units (left column) and 5 D L I units (right column) from one subject. Fig. 5 shows histograms from 5 M E N T units (left column) and 5 D L I units (right column) in a different subject. For both subjects the magnitudes of the suppression responses in the D L I units are larger than in the elevator units. This tendency for D L I units to show larger S responses than O O I and M E N T units is also indicated in Table II.
71
Magnitudes of the reflex responses The magnitudes of the various reflex responses, as measured in terms of additional pulses per stimulus, were somewhat larger in the lip muscles than those reported for single motor units in the first dorsal interosseous muscle by Garnett artd Stephens t~. They reported a range of---0.1-0.3 additional pulses per stimulus, whereas 18 % of the responses in the present study exceeded this range. However, differences in the nature Qf the stimulus must be taken into account when interpreting this result. Consistent with the findings of Garnett and Stephens, there was substantial intersubject variability in the magnitude of responses in the present study. For example, the magnitude of E1 ranged from a maximum in one subject of 0.11 additional pulses per stimulus to a maximum of 0.88 in another subject. In order to determine whether the size of the reflexes studied here was dependent on stimulus magnitude, 9 motor units sampled from 5 subjects
wore tested at increasing levels of mechanical stimulation. The 10 ms tap stimulus was varied in steps of 0.1 mm between 0.4 and 0.9 mm, and 512 stimuli were delivered at each of the 6 levels. The results of this test on an OOI motor unit are shown in Fig. 6 which contains the peristimulus histograms obtained at each of the 6 levels of stimulation. The size of the E 1 response increased with increases in stimulus magnitude. Of the 9 motor units tested, 8 had at least one response component which displayed a significant positive correlation (P < 0.05) with stimulus magnitude, and 11 of the 16 responses present had significant positive correlations with stimulus magnitude. Significant positive correlations were seen for El, E2 and S responses. It has boon reported that the second component of the perioral reflex (E2) tends to habituate with repeated stimulation n. To test if the excitatory reflex responses described in the present study habituated with repeated stimulation, 13 motor units
)
Fig. 4. Peristimulus time histograms obtained on 5 0 0 I motor units (on the left) and 5 DLI motor units (on the right) in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus.
72
Fig. 5. Peristimulus time histograms obtained on 5 MENT units (on the left) and 5 DLI units (on the right) in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus. sampled from 5 subjects were further analyzed. For these units the magnitudes of E1 and E2 were compared across 3 consecutive sets of 256 trials. There was no indication that repeated stimulation had a habituating effect on either El or E2.
The effects of stimulus location on 0 0 I motor unit responses The relatively high percentage of El response patterns in the OOI motor units and the fact that the mechanical stimulus was applied at a location closest to the OOI muscle suggests that the E1 response may represent a reflex in which the primary afferents mediating t h e response are most sensitive to stimulation in the region of the muscle itself. In order to investigate this possibility seven OOI motor units recorded from 4 subjects were tested with the standard 0.5 m m tap stimulus applied both to the corner of the mouth and the chin. The stimulus was applied to the chin between the corner of the mouth and the midsagittal plane and between the corner of
the mouth and the inferior border of the mandible. The stimulation probe was positioned approximately 75 ~ of the distance medially and inferiorly from the corner of the mouth. Six of the 7 OOI units tested showed an E1 response to stimulation at the corner of the mouth, but none of the 7 showed an El response to chin stimulation. S and E2 responses were typically present for both stimulus locations.
The effects o f mechanical stimulation of the teeth With stimulation of oral-facial structures it is often difficult to argue that only one structure is stimulated. In the present experiment it was problematic that the tap might be translated through the soft tissues of the lips to the teeth. Tooth mechanoreceptors are known to have powerful reflex effects on jaw muscles 8,7, but their possible reflex effects on human lip muscles have not been explored. In order to provide evidence that tooth mechanoreceptors were not the major contributor to the
73 reflex responses described above, we investigated the effects of mechanical stimulation of the teeth on lower lip motor unit discharge. Twelve motor units ( 3 0 0 I , 3 DLI, 6 MENT) were tested in 4 subjects with mechanical taps applied to the right mandibular lateral incisors with precaution taken not to stimulate the lips. This was accomplished by having the subject hold a small rubber cylinder between the lips through which the stimulation probe could be positioned against the teeth. Stimulation of the lips was done with this cylinder in place. When stimula-
ting the teeth the tap displacement was reduced in order that the peak force applied to the teeth and lips was approximately equivalent. This force was between 5 and 10 g. The 12 motor units showed 23 responses to lip stimulation and 5 responses to tooth stimulation. These 5 responses (4 E2 and one S) occurred in 4 units, and they were all smaller in magnitude (less than half in 4 of 5 cases) than the corresponding response to lip stimulation. On the basis of these findings it appears reasonable to conclude that the reflex responses observed with stimulation at the corner of the mouth were not primarily the result of indirect stimulation of the underlying teeth. If there was some contribution from indirect tooth stimulation, it was a small effect on E2 or S responses. DISCUSSION
Fig. 6. Peristimulus histograms obtained on the same OOI motor unit using 6 levels of stimulation. The level of stimulus displacement increased in 0.1 ram steps from 0.4--0.9mm progressing from the top to the bottom histogram. In each case 512 stimuli were applied. The calibration bars are 25 counts and 25 ms. The vertical line through the histograms indicates the time of onset of the stimulus.
Small mechanical stretches applied at the corner of the mouth were found to produce changes in the probability of firing of single motor units recorded from 3 lower lip muscles. The distribution of latenties of the excitatory responses were consistent with previous descriptions of the perioral reflexesS,11, 20. A novel finding was that mechanical stimulation frequently produced suppression responses in lower lip motor units. There is some previous evidence of reflex suppression in the human perioral muscles in response to mechanical stimulation2Z; however, it has been reported that reflex suppression occurs rarely in the facial muscles when compared with spinal muscles24, 25. It seems probable that earlier studies failed to reveal suppression responses in the perioral muscles because their methods did not allow the detection of decreases in E M G activityS, 11 or the differentiation of DLI responses from those of other muscles 20. As to the origin of the suppression responses described in the present study, some were probably due to motoneuron refractoriness, since they occurred after large excitatory responses. However, in many cases the suppression responses occurred initially or their magnitude far exceeded that of a prior excitatory effect. Such suppressions could have resulted from disfacilitation of tonic excitatory inputs to the facial nucleus or excitation of inhibitory interneurons within the trigeminal complex. The latter possibility is supported by anatomical evi-
74 dence for inhibitory interneurons in the spinal trigeminal nucleus13, and the occurrence of IPSP's in cat facial nucleus motoneurons following trigeminal nerve stimulation21. Mechanical stimulation tended to produce different reflex response patterns in the motor units associated with the 3 lip muscles, although there was some degree of overlap in response patterns. E1 occurred most often and tended to be the largest reflex response in OOI motor units. Since the OOI units were closest to the site of mechanical stimulation and the El response was not present when the stimulus was applied to the chin, E1 responses may depend upon activation of receptors lying in the skin over the sampled muscle or upon activation of other receptors closely associated with the muscle itself. E2 responses occurred with equal frequency among the 3 muscles, but constituted the largest response in 72~ of MENT units and 51 ~ of DL1 units. The major difference between the MENT and DLI unit response patterns was that the number and magnitude of S responses was greater in DLI units than in MENT units. In 45 ~ of the DLI units, S was the largest response component; while in 7 ~ of the MENT units, S was the largest response. The present results provide preliminary evidence that in humans the trigeminal projections to facial nucleus motoneuron pools are organized with some degree of muscle specificity. That muscle-specific trigeminofacial projections do exist is also suggested by neuroanatomical data in the rat. Erzurumlu and Killackeyl° found that the subnucleus caudalis, a region of the spinal trigeminal tract mediating sensory input from the vibrissae, projects exclusively within the facial nucleus to the lateral cell group which contains the motoneurons innervating the muscles controlling vibrissae movement. The overlap in response categories in Tables I and II may have resulted in part from inappropriate muscle classification of some motor units. The overlap also may have resulted from excitation of trigeminal afferents having different degrees of musclespecific organization in their projection to the facial nucleus; since the displacement stimulus could have activated a variety of receptors including encapsulated and free nerve endings, and typical and atypical muscle spindles ~,17,19. It has been reported that the effects of cutaneous
stimulation are different for low and high threshold motor units within the human first dorsal interosseous muscle 1'. Two observations suggest that our methods tended to select primarily low threshold motor units for study. The preponderance of motor units selected for testing were activated by slight muscle contractions for the simple reason that subiects had difficulty maintaining forceful isometric contractions of these muscles. In addition, we isolated a few motor units that could be activated only in rapid movements. These units could not be tested under our procedure which required lengthy tonic discharge of the motor units. Thus, the reflex effects described in the present study may pertain primarily to low threshold motor units. However, since the histochemical features of the perioral muscles in humans are unknown and their unusual geometry prevents measurement of twitch tensions by signal averaging techniques used in other human muscles 23, it is not certain that the lip muscles are heterogeneous with respect to motor unit type. It is apparent from the above discussion that many of the basic physiological features of the human perioral muscles are unknown. The present description of the perioral reflexes in terms of responses in single motor units suggests that although the territory of individual muscles is overlapping, the muscles are not entirely undifferentiated in their reflex responses to mechanical stimulation. An understanding of the functional significance of this organization will require more detailed information on the properties of perioral muscles and mechanoreceptors, as well as the development of improved techniques for studying the perioral reflexes during behaviors such as speech production and mastication. ACKNOWLEDGEMENTS This research was supported by the National Institute of Neurological and Communicative Disorders and Stroke (NS 14048, NS 00291) and the University of Washington Biomedical Research Support Grant. We express our thanks to Erich Luschei and Charles Larson for their useful comments on an earlier version of this manuscript. It was judged that the experimental procedures used in the present study did not constitute a significant
75 health hazard to the subjects. The use of human subjects irt this experiment was approved by the Univer-
sity
REFERENCES
scand., 26 (1952) Suppl. 94. 15 Hagbarth, K., Spinal withdrawal reflex in human lower limbs, J. Neurol. Neurosurg. Psychiat., 23 (1960) 222227. 16 Hollinshead, W., Anatomy for Surgeons: The Head and Neck, Vol. 1, Harper and Row, New York, 1968. 17 Hosokawa, H., Proprioceptive innervation of striated muscles in the territory of cranial nerves, Tex. Rep. Biol. Med., 19 (1961) 405-464. 18 Huntsman, L., Nichols, G., Verdugo, P. and Pollack, G., A new instrument for mechanical testing of tissues, Proc. Ann. Conf. Eng. Med. Biol., 24 (1971) 50. 19 Kadanoff, D., Die sensiblen narvendigungen in der mimshen muskulatur des menschen, Z. mikr.-anat. Forsch., 62 (1956) 1-15. 20 Larson, C., Folkins, J., McClean, M. and Muller, E., Sensitivity of the human perioral reflex to parameters of mechanical stimulation, Brain Research, 146 (1978) 159-164. 21 Limanskii, Y., Pilyavskii, A. and Gura, E., Postsynaptic potentials of facial nucleus motoneurons evoked by afferent and corticofugal impulses, Neurophysiology, 5 (1973) 296-303. 22 McClean, M. and Cooker, H., Preliminary observations on the response of perioral single motor units to mechanical stimulation, J. acoust. Soc. Amer, 67 (1980) 271-275. 23 Milner-Brown, H., Stein, R. and Yemm, R., The contractile properties of human motor units during voluntary isometric contractions, J. Physiol. (Lond.), 228 (1973) 285-306. 24 Penders, C. and Delwaide, P., Physiologic approach to the human blink reflex. In J. Desmedt (Ed.), New Developments in Electromyography and Clinical Neurophysiology, VoL 3, Karger, Basel, 1973, pp. 649-657. 25 Shahani, B. and Young, R., Studies of the normal human silent period. In J. Desmedt (Ed.), New Develepments in Electromyography and Clinical Neurophysiology, VoL 3, Karger, Basel, 1973, pp. 641-648. 26 Smith, A., Zimrnermann, G. and Abbas, P., Recruitment of motor units in speech production, J. Speech Res., 24 (1981) 567-575. 27 Stephens, J., Usherwood, T. and Garnett, R., Technique for studying synaptic connections of single motoneurons in man, Nature (Lond.), 263 (1976) 343-344.
1 Ashby, P. and Labelle, K., The effects of extensor and flexor group I afferent volleys on the excitability of individual soleus motoneurons in man, J. Neurol. Neurosurg. Psychiat., 40 (1977) 910-919. 2 Bak, M. and Sehmidt, E., An improved time-amplitude window discriminator, IEEE Trans. Biomed. Eng., 24 (1977) 486--489. 3 Bratzlavsky, M., Human bralnstem reflexes. In M. Shahani (Ed.), The Motor System: Neurophysiology and Muscle Mechanisms, Elsevier, Amsterdam, 1976, pp. 135-154. 4 Buller, N., Garnett, R. and Stephens, J., The reflex responses of single motor units in human hand muscles following muscle afferent stimulation, J. Physiol. (Lond.), 303 (1980) 337-340. 5 Caccia, M., McComas, A., Upton, A. and Blogg, T., Cutaneous reflexes in small muscles of the hand, J. NeuroL Neurosurg. Psychiat., 36 (1973) 960-972. 6 Darian-Smith, I., Neural mechanisms of facial sensation, Int. Rev. Neurobiol., 9 (1973) 301-395. 7 Dubner, R., Sessle, B. and Storey, A., The Neural Basis of Oral and Facial Function, Plenum Press, New York 1978, 483 pp. 8 Ekbom, K., Jernelius, M. and Kugelberg, E., Perioral reflexes, Neurology, 2 (1952) 103-111. 9 Ellaway, P., Cumulative sum technique and its application to the analysis of peristimulus time histograms, Electroenceph. Clin. Neurophysiol., 45 (1978) 302-304. 10 Erzurumlu, R. and Killackey, H., Efferent connections of the brainstem trigaminal complex with the facial nucleus of the rat, J. comp. NeuroL, 188 (1979) 75-86. 11 Gandiglio, C. and Fra, L., Further observations on facial reflexes, J. neuroL Sci., 5 (1967) 173-185. 12 Garnett, R. and Stephens, J., The reflex response of single motor units in human first dorsal interosseous muscle following cutaneous afferent stimulation, J. Physiol. (Lond.), 303 (1980) 351-364. 13 Gobel, S., Some considerations of synaptic organization in the trigeminal sensory nuclei of the adult cat. In B. Sessle and A. Hannam (Eds.), Mastication and Swallowing: Biological and Clinical Correlates, University Toronto Press, Toronto, 1976, pp. 351-364. 14 Hagbarth, K., Excitatory and inhibitory skin areas for flexor and extensor motoneurones, Acta physiol.
of
Washington
Human
Subjects
Review
Committee.
65
The Reflex Responses of Single Motor Units in Human Lower Lip Muscles to Mechanical Stimulation MICHAEL D. McCLEAN and ANNE SMITH Department of Speech and Hearing Sciences and ( A.S.) Department of Physiology and Biophysics, University of Washington, Seattle, WA 98195 (U.S.A.)
(Accepted April 22nd, 1982) Key words: reflexes - - single motor units - - human lower lip
Single motor unit activity was recorded from human lower lip muscles while subjects maintained a firing rate of 20 ips using audio and visual feedback. Each motor unit was classified as belonging to one of three muscles, orbicularis oris inferior (OOI), mentalis (MENT), or depressor labii inferior (DLI). Reflex responses were elicited in single motor units by applying small mechanical displacements (0.5 mm) to the corner of the mouth. Reflex responses were analyzed by constructing peristimulus histograms. Of 130 units analyzed, 123 showed statistically significant changes in their probability of firing in the 50 ms period following the stimulus. Reflex responses were typically polymodal with both excitatory and suppression components present. The excitatory responses occurred at latencies consistent with previous descriptions of the perioral reflexes and were labelled El (X = 15.9 ms) and E2 (X = 35.6 ms). Reflex suppression responses (S) were also consistently observed. There was a tendency for the pattern and/or relative magnitudes of reflex response components to depend on the muscle classification of the motor unit; E1 responses predominating in OOI units, E2 responses in MENT units, and S responses in DLI units. This finding provides preliminary evidence for muscle-specificprojections between the trigeminal sensory system and facial nucleus motoneurones in humans. INTRODUCTION Electrical stimulation of trigeminal nerve branches or mechanical stimulation of the perioral tissues produces diffuse excitatory reflex responses in human lip musclesS,11, 20. Little is known, however, about the specific organization of trigeminal inputs to facial nucleus motoneurons, and there is no clear evidence that the reflex effects of mechanical stimulation are different for different perioral muscles. In order to study the perioral reflexes with quantitative mechanical stimuli and to determine whether or not there are differing patterns of effects on motoneurons innervating closely adjacent but functionally distinct muscles, we have adopted the methods used by Stephens and his colleagues4,12, zT. In this approach, subjects use visual and/or auditory feedback to maintain a constant firing rate in a motor unit, and the effects of repeated stimulation are determined by constructing a poststimulus time histogram. Changes in the probability of firing timelocked to the stimulus reflect the contribution made 0006-8993/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press
by the stimulus to the total firing pattern of the cell. Stephens and colleagues4,12, 27 have shown that innocuous mechanical stimuli produce reflex effects on motoneurons of human hand muscles. These responses were large and elicited only by skin stimulation in areas whose natural stimulation could be expected to be associated with movements involving activity in a particular muscle. Garnett and Stephens (ref. 12) suggested, therefore, that cutaneous input may play a major role in modifying motor unit discharge during movement. The analysis of reflex effects on single motor unit activity is important to the study of lip muscles because the fibers of different lip muscles are often interlaced. For example, the fibers of antagonist muscles, mentalis (a lower lip elevator) and depressor labii inferior (a lower lip depressor), blend at their common insertion on the skin of the chin in. Thus, the activity of one muscle cannot be determined with certainty from interference-pattern E M G recordings. Smith et al. an reported that even a fine-wire single motor unit electrode with a distance
66 between leads of less than 25 # m typically recorded both lower lip elevator and depressor units at a single intramuscular site in the chin. Previous descriptions of the perioral reflexes and the fact that the lip muscles are not spatially separate suggest that trigeminal inputs to the facial nucleus may not be organized with respect to individual muscles. However, on the basis of studies of human and cat spinal muscle systems1,5,12,14,15 and data on trigeminofacial connections in the rat 10, it seems possible that a muscle-specific organization does exist. The present experiment was undertaken to investigate the reflex effects of mechanical stimulation on single motor units in human lip muscles having different voluntary functions and different locations relative to the point of stimulation. The discovery of any systematic relationships between lower lip motor unit function, location, and pattern of reflex response should provide useful information on the organization of trigeminal input to the facial nucleus in humans as well as a more refined basis for experiments designed to investigate the role of the perioral reflexes in the control of oral-facial movements.
Subjects
muscle that surrounds the lower lip and functions to elevate and round the lower lip. 2. Mentalis (MENT), a muscle that arises from the mandible, inserts in the skin of the chin and functions to elevate the lower lip. 3. Depressor labii inferior (DL1), a muscle that arises from the mandible, inserts in the skin of the chin and the skin and mucosa of the lower lip, and functions to lower the lower lip. The location of each motor unit (the estimated point at which the needle ended after the insertion) was placed within a Cartesian coordinate system in which the right corner of the mouth was taken as the origin. Vertical and horizontal measurements were converted to percentage values based on the facial dimensions of the subject. This procedure allowed motor units from different subjects to be compared in terms of their relative locations. Each motor unit was classified as either a lower lip elevator or depressor on the basis of (1) its activation characteristics during isotonic speech and nonspeech movements of the lower lip as observed on a storage oscilloscope, and (2) the posturing of the lips during isometric activation associated with motor unit training. In a very few cases motor units did not clearly behave like a lower lip elevator or depressor on both of these types of gestures. These units were excluded from further analysis.
Subjects were seven adult volunteers. All were apparently healthy; one was a mild stutterer.
Muscle class(fication of motor units
MATERIALS AND METHODS
Single motor unit recording Single motor units were recorded with bipolar hooked-wire electrodes. The electrodes were made of two Teflon-insulated stainless steel wires (25 # m coated diameter) twisted tightly together. The ends of the wires were burned and the exposed tips were then trimmed back until the impedance of each wire was 30-300 kOhms measured at I kHz in saline with a Wheatstone bridge. The distance between the two leads was less than 25 #m. The electrodes were inserted into the muscle with a 30-gauge hypodermic needle which was then withdrawn. Motor unit potentials were amplified with a high impedance differential amplifier (Honeywell, Model 135). Art attempt was made to record single motor unit potentials from 3 lip muscles. 1. Orbicularis oris inferior (OOI), a sphincter-like
The close proximity and complex anatomy of the human perioral muscles make it difficult to classify single motor unit action potentials with particular muscles on the basis of electrode location. For example, DLI is known to blend with fibers from OOI and M E N T at the regions of its insertion into the lower lip and chin. Consistent with this observation is the earlier mentioned finding that both lower lip elevator and depressor motor units have been recorded over the same fine-wire electrode configuration in the chin z6. At our present stage of understanding of the perioral motor system, it seems most parsimonious to assume that such motor units represent distinct muscles rather than motor unit 'subpopulations' of one muscle. For the above reasons, strictly functional criteria were used in the present study to distinguish DLI from OOI and M E N T motor units. All motor units
67 which ftmctioned to depress the lower lip were classified as DLI units, regardless of electrode location. All motor units which functioned to elevate the lower lip were classified as either OOI or M E N T units. The OOI and M E N T motor units were distinguished from one another solely on the basis of anatomical criteria, since these muscles have a shared function (lower lip elevation) but relatively distinct anatomical locations. Classification of individual OOI and M E N T motor units was made by comparing the electrode location with previous descriptions of the anatomy of these muscles.
Mechanical stimulation Small mechanical stretches were applied to the right corner of the subject's mouth with a positioncontrolled electromagnetic servo systemlS, 20. A force transducer with a rubber-tipped cylindrical extension (diameter = 5 mm) was fixed to the end of the displacement shaft. The rubber-tipped cylinder was placed in contact with the skin at the corner of the mouth at a standing force of 20 g. The standard stimulus was a tap with a duration of 10 ms, a rise time of 4 ms, and a displacement of 0.5 mm. Such stimuli were applied at a rate of between 3 and 4 taps/s. A pre-stimulus trigger signal occurring 50 ms before the tap, a voltage analog of the shaft position, and the E M G signal were recorded on a Honeywell FM tape recorder (Model 5600). The frequency response in all 3 channels was flat from DC to 2.5 kHz.
Experimental procedure Subjects were seated in a dental chair with their heads supported by a comfortable headrest. The single motor unit signal was fed back by means of a loudspeaker and an oscilloscope to aid subjects in controlling motor unit firing rate. The oscilloscope was triggered on the motor unit potential and the time base set so that subjects could maintain a rate of 20 ips. After the subject had practiced long enough to maintain relatively steady rates of discharge, stimuli were delivered to the corner of the mouth. The force signal was displayed on one channel of the oscilloscope to aid in the maintenance of the standing force which varied with 'drift' in head position. Motor unit activation and stimulation were sustained for a period of at least 3 min in
order that a minimum of 512 stimuli could be applied.
Data analysis Peristimulus time histograms were constructed for each motor unit with an Ortcc Time Histogram Analyzer (Model 4620A-4621). The bin width was 1 ms, and a 255 ms period was analyzed including a 50 ms pre-stimulus interval. During the generation of a histogram each motor unit was monitored on an oscilloscope to insure that only one unit triggered the computer. In some cases, when more than one unit was active, a Time-Amplitude Window Discriminator (Bak Electronics, Model DIS-1) was used to isolate smaller amplitude motor unit potentials 2. The contents of the Ortec memory were read into a Tektronix 4051 computer for signal processing, measurement, and statistical analysis. Reflex responses were initially identified by inspecting the peristimulus histograms on the computer terminal screen. A cumulative sum plot 9 was displayed below each histogram, and it was used to make a more precise determination of the temporal onset and offset of individual responses. An interactive computer program was used to impose time markers at the points of visually detectable continuous changes in the slope of the cumulative sum plot. The computer then automatically calculated the latency, duration, magnitude and statistical significance of each response. The magnitude of reflex responses was quantified as additional pulses per stimulus. This number was calculated by subtracting the prestimulus control level from the total number of spikes occurring over the duration of the response and dividing by the number of stimuli. The statistical significance of individual responses was assessed with Poisson statistics as described elsewhere 12. Only responses reaching significance at the 5 ~ level in a two-tailed test are reported in the present study. RESULTS Histograms were constructed on a total of 130 single motor units: 38 OOI units, 43 M E N T units and 49 DLI units. Fig. 1 shows the estimated locations of the motor units. Of the 130 units, 123 showed a significant change in probability of firing in the 50 ms period following the onset of the
68
O
MENT
•
DLI
8
001
Fig. 1. An illustration of the estimated locations of 122 motor units analyzed in the present study. stimulus. Significant responses sometimes occurred at poststimttlus latencies longer than 50 ms; however, these longer latency responses are not considered in the present report. Fig. 2 illustrates for an OO1 unit the graphic display used to measure the temporal onset and offset of the component responses for each motor unit, The top trace is a peristimulus time histogram, and the bottom trace is the cumulative sum plot calculated on this histogram. The mean firing rate of the motor units studied, as estimated from the bin counts in the 50 ms prestimulus interval, was 18.4 impulses/s (S.D. = 4.4).
Classification of motor unit responses A wide variety of polymodal response patterns was observed in the histograms of the various motor units studied. In order to describe these effects, a system was developed for assigning each motor unit to a unique class based on the temporal characteristics of the excitatory artd suppression phases in its peristirnulus histogram. The reflex responses of the
lip muscles have been described as having two excitatory components with latertcies of 10-17 ms and 25-45 ms s,11,2°. The poststimulus latencies of the excitatory responses observed in the present study were consistent with this description. These latencies were bimodally distributed with a first component response (El) having a range of 8-21 ms ('~ = 15.9, S.D. ~- 3.8, n ~ 51) and a second component response (E2) having a range of 21-50 ms ( ~ = 35.6, S.D. = 7.2, n ---- 105). The latencies of the suppression responses (S) obtained in the present study were unimodally distributed with a range of 13-50 ms (.~ ---- 28.2, S.D. -~ 8.3, n = 98). On the basis of these latency distributions, the excitatory and suppression responses in each histogram were classified as El, E2 or S responses. Eighteen motor unit response classes were found; however, 10 of these classes contained only one or two motor units and together accounted for less than 10 ~ of all the units. In order to simplify the classification system, response classes containing only one or two motor
69 1BB
f PSTH
C
N 50 T
S
t
s
E P R
I
D HE
EMcS 201 -20e .
'
"
.
.
.
"
,
e
.
.
•
.
,
.
.
~e
.
.
.
.
.
.
,
TIME(MSEC) .
.
.
.
tee
.
.
,
.
.
.
.
ISe
.
.
,
.
•
.
.
.
.
2~e
Fig. 2. A peristimulus time histogram (PSTH) and cumulativesum plot (CUMSUM) on an OOI motor unit activated at an average firing rate of 26 impulses/s. The 3 types of response components identified in the present study are indicated by the labels El, E2 and S. See the text for further explanation of this classification. units were combined with other response classes. This involved, for example, combining a category such as E1-S-E2-S with El-S-E2. Table I shows the classification of response pat-
terns of the motor units recorded from the 3 muscles. The number in each cell is the percentage of motor units within a given muscle class that showed the associated response pattern.
TABLE I
R e f l e x responses in lip elevator m o t o r units
Percentages of 001, M E N T and DLI motor units showing the various histogram response patterns
See text for explanation of the response-pattern categories. NR indicates that no significant responses occurred in the 50 ms poststimulus period.
El-E2 EI-S El-S-E2 E2 E2-S S-E2 S NR
00I (n = 38)
MENT (n = 43)
DLI (n = 49)
39 26 16 13 0 0 3 3
9 7 9 21 7 33 2 12
6 8 0 6 24 41 12 2
Table I indicates that differences exist in the reflex response patterns of the two groups of elevator motor units, OOI and MENT. The percentage of OOI units showing E1 response patterns was greater than the percentage of M E N T units with E1 response patterns. The M E N T units, on the other hand, showed a greater percentage of E2, S-E2, and E2-S patterns compared to the OOI units. This difference in the response patterns of OOI and M E N T motor units is further illustrated in Fig. 3 which shows peristimulus histograms obtained on 5 OOI units (left column) and 5 M E N T units (right column) in the same subject. It may also be seen in Fig. 3 that the OOI units had larger El responses compared with the E2
70
Fig. 3. Peristimulus time histograms obtained on 5 0 0 I motor units (on the left) and 5 MENT units (on the righ0 in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus. responses, and the M E N T units had larger E2 responses compared with the E1 responses. This difference represents a general trend seen for all the O O I and M E N T units, and it is further illustrated in Table II. Table II shows the percentage of m o t o r units in each muscle class in which the largest reflex response was El, E2 or S. TABLE II The percentage o f motor units in each muscle class in which the largest reflex response was El, E2, or S N R indicates that no significant 50 ms poststimulus time interval.
E1 E2 S NR
responses occurred in the
001 (n : 38)
MENT (n = 43)
DLI (n = 49)
63 29 5 3
9 72 7 12
2 51 45 2
R e f l e x r e s p o n s e s in lip d e p r e s s o r m o t o r units
Table I indicates that the D L I motor units had a greater percentage of E2-S, S-E2, and S response patterns compared to the elevator units (OOI and MENT). As with the M E N T units, the D L I units showed few E1 response patterns relative to the OOI units. The differences in the reflex responses of the depressor and elevator motor units are further illustrated in Figs. 4 and 5. Fig. 4 shows the peristimulus histograms of 5 OOI units (left column) and 5 D L I units (right column) from one subject. Fig. 5 shows histograms from 5 M E N T units (left column) and 5 D L I units (right column) in a different subject. For both subjects the magnitudes of the suppression responses in the D L I units are larger than in the elevator units. This tendency for D L I units to show larger S responses than O O I and M E N T units is also indicated in Table II.
71
Magnitudes of the reflex responses The magnitudes of the various reflex responses, as measured in terms of additional pulses per stimulus, were somewhat larger in the lip muscles than those reported for single motor units in the first dorsal interosseous muscle by Garnett artd Stephens t~. They reported a range of---0.1-0.3 additional pulses per stimulus, whereas 18 % of the responses in the present study exceeded this range. However, differences in the nature Qf the stimulus must be taken into account when interpreting this result. Consistent with the findings of Garnett and Stephens, there was substantial intersubject variability in the magnitude of responses in the present study. For example, the magnitude of E1 ranged from a maximum in one subject of 0.11 additional pulses per stimulus to a maximum of 0.88 in another subject. In order to determine whether the size of the reflexes studied here was dependent on stimulus magnitude, 9 motor units sampled from 5 subjects
wore tested at increasing levels of mechanical stimulation. The 10 ms tap stimulus was varied in steps of 0.1 mm between 0.4 and 0.9 mm, and 512 stimuli were delivered at each of the 6 levels. The results of this test on an OOI motor unit are shown in Fig. 6 which contains the peristimulus histograms obtained at each of the 6 levels of stimulation. The size of the E 1 response increased with increases in stimulus magnitude. Of the 9 motor units tested, 8 had at least one response component which displayed a significant positive correlation (P < 0.05) with stimulus magnitude, and 11 of the 16 responses present had significant positive correlations with stimulus magnitude. Significant positive correlations were seen for El, E2 and S responses. It has boon reported that the second component of the perioral reflex (E2) tends to habituate with repeated stimulation n. To test if the excitatory reflex responses described in the present study habituated with repeated stimulation, 13 motor units
)
Fig. 4. Peristimulus time histograms obtained on 5 0 0 I motor units (on the left) and 5 DLI motor units (on the right) in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus.
72
Fig. 5. Peristimulus time histograms obtained on 5 MENT units (on the left) and 5 DLI units (on the right) in the same subject. The calibration bars are 25 counts and 25 ms. The vertical lines through the histograms indicate the time of onset of the stimulus. sampled from 5 subjects were further analyzed. For these units the magnitudes of E1 and E2 were compared across 3 consecutive sets of 256 trials. There was no indication that repeated stimulation had a habituating effect on either El or E2.
The effects of stimulus location on 0 0 I motor unit responses The relatively high percentage of El response patterns in the OOI motor units and the fact that the mechanical stimulus was applied at a location closest to the OOI muscle suggests that the E1 response may represent a reflex in which the primary afferents mediating t h e response are most sensitive to stimulation in the region of the muscle itself. In order to investigate this possibility seven OOI motor units recorded from 4 subjects were tested with the standard 0.5 m m tap stimulus applied both to the corner of the mouth and the chin. The stimulus was applied to the chin between the corner of the mouth and the midsagittal plane and between the corner of
the mouth and the inferior border of the mandible. The stimulation probe was positioned approximately 75 ~ of the distance medially and inferiorly from the corner of the mouth. Six of the 7 OOI units tested showed an E1 response to stimulation at the corner of the mouth, but none of the 7 showed an El response to chin stimulation. S and E2 responses were typically present for both stimulus locations.
The effects o f mechanical stimulation of the teeth With stimulation of oral-facial structures it is often difficult to argue that only one structure is stimulated. In the present experiment it was problematic that the tap might be translated through the soft tissues of the lips to the teeth. Tooth mechanoreceptors are known to have powerful reflex effects on jaw muscles 8,7, but their possible reflex effects on human lip muscles have not been explored. In order to provide evidence that tooth mechanoreceptors were not the major contributor to the
73 reflex responses described above, we investigated the effects of mechanical stimulation of the teeth on lower lip motor unit discharge. Twelve motor units ( 3 0 0 I , 3 DLI, 6 MENT) were tested in 4 subjects with mechanical taps applied to the right mandibular lateral incisors with precaution taken not to stimulate the lips. This was accomplished by having the subject hold a small rubber cylinder between the lips through which the stimulation probe could be positioned against the teeth. Stimulation of the lips was done with this cylinder in place. When stimula-
ting the teeth the tap displacement was reduced in order that the peak force applied to the teeth and lips was approximately equivalent. This force was between 5 and 10 g. The 12 motor units showed 23 responses to lip stimulation and 5 responses to tooth stimulation. These 5 responses (4 E2 and one S) occurred in 4 units, and they were all smaller in magnitude (less than half in 4 of 5 cases) than the corresponding response to lip stimulation. On the basis of these findings it appears reasonable to conclude that the reflex responses observed with stimulation at the corner of the mouth were not primarily the result of indirect stimulation of the underlying teeth. If there was some contribution from indirect tooth stimulation, it was a small effect on E2 or S responses. DISCUSSION
Fig. 6. Peristimulus histograms obtained on the same OOI motor unit using 6 levels of stimulation. The level of stimulus displacement increased in 0.1 ram steps from 0.4--0.9mm progressing from the top to the bottom histogram. In each case 512 stimuli were applied. The calibration bars are 25 counts and 25 ms. The vertical line through the histograms indicates the time of onset of the stimulus.
Small mechanical stretches applied at the corner of the mouth were found to produce changes in the probability of firing of single motor units recorded from 3 lower lip muscles. The distribution of latenties of the excitatory responses were consistent with previous descriptions of the perioral reflexesS,11, 20. A novel finding was that mechanical stimulation frequently produced suppression responses in lower lip motor units. There is some previous evidence of reflex suppression in the human perioral muscles in response to mechanical stimulation2Z; however, it has been reported that reflex suppression occurs rarely in the facial muscles when compared with spinal muscles24, 25. It seems probable that earlier studies failed to reveal suppression responses in the perioral muscles because their methods did not allow the detection of decreases in E M G activityS, 11 or the differentiation of DLI responses from those of other muscles 20. As to the origin of the suppression responses described in the present study, some were probably due to motoneuron refractoriness, since they occurred after large excitatory responses. However, in many cases the suppression responses occurred initially or their magnitude far exceeded that of a prior excitatory effect. Such suppressions could have resulted from disfacilitation of tonic excitatory inputs to the facial nucleus or excitation of inhibitory interneurons within the trigeminal complex. The latter possibility is supported by anatomical evi-
74 dence for inhibitory interneurons in the spinal trigeminal nucleus13, and the occurrence of IPSP's in cat facial nucleus motoneurons following trigeminal nerve stimulation21. Mechanical stimulation tended to produce different reflex response patterns in the motor units associated with the 3 lip muscles, although there was some degree of overlap in response patterns. E1 occurred most often and tended to be the largest reflex response in OOI motor units. Since the OOI units were closest to the site of mechanical stimulation and the El response was not present when the stimulus was applied to the chin, E1 responses may depend upon activation of receptors lying in the skin over the sampled muscle or upon activation of other receptors closely associated with the muscle itself. E2 responses occurred with equal frequency among the 3 muscles, but constituted the largest response in 72~ of MENT units and 51 ~ of DL1 units. The major difference between the MENT and DLI unit response patterns was that the number and magnitude of S responses was greater in DLI units than in MENT units. In 45 ~ of the DLI units, S was the largest response component; while in 7 ~ of the MENT units, S was the largest response. The present results provide preliminary evidence that in humans the trigeminal projections to facial nucleus motoneuron pools are organized with some degree of muscle specificity. That muscle-specific trigeminofacial projections do exist is also suggested by neuroanatomical data in the rat. Erzurumlu and Killackeyl° found that the subnucleus caudalis, a region of the spinal trigeminal tract mediating sensory input from the vibrissae, projects exclusively within the facial nucleus to the lateral cell group which contains the motoneurons innervating the muscles controlling vibrissae movement. The overlap in response categories in Tables I and II may have resulted in part from inappropriate muscle classification of some motor units. The overlap also may have resulted from excitation of trigeminal afferents having different degrees of musclespecific organization in their projection to the facial nucleus; since the displacement stimulus could have activated a variety of receptors including encapsulated and free nerve endings, and typical and atypical muscle spindles ~,17,19. It has been reported that the effects of cutaneous
stimulation are different for low and high threshold motor units within the human first dorsal interosseous muscle 1'. Two observations suggest that our methods tended to select primarily low threshold motor units for study. The preponderance of motor units selected for testing were activated by slight muscle contractions for the simple reason that subiects had difficulty maintaining forceful isometric contractions of these muscles. In addition, we isolated a few motor units that could be activated only in rapid movements. These units could not be tested under our procedure which required lengthy tonic discharge of the motor units. Thus, the reflex effects described in the present study may pertain primarily to low threshold motor units. However, since the histochemical features of the perioral muscles in humans are unknown and their unusual geometry prevents measurement of twitch tensions by signal averaging techniques used in other human muscles 23, it is not certain that the lip muscles are heterogeneous with respect to motor unit type. It is apparent from the above discussion that many of the basic physiological features of the human perioral muscles are unknown. The present description of the perioral reflexes in terms of responses in single motor units suggests that although the territory of individual muscles is overlapping, the muscles are not entirely undifferentiated in their reflex responses to mechanical stimulation. An understanding of the functional significance of this organization will require more detailed information on the properties of perioral muscles and mechanoreceptors, as well as the development of improved techniques for studying the perioral reflexes during behaviors such as speech production and mastication. ACKNOWLEDGEMENTS This research was supported by the National Institute of Neurological and Communicative Disorders and Stroke (NS 14048, NS 00291) and the University of Washington Biomedical Research Support Grant. We express our thanks to Erich Luschei and Charles Larson for their useful comments on an earlier version of this manuscript. It was judged that the experimental procedures used in the present study did not constitute a significant
75 health hazard to the subjects. The use of human subjects irt this experiment was approved by the Univer-
sity
REFERENCES
scand., 26 (1952) Suppl. 94. 15 Hagbarth, K., Spinal withdrawal reflex in human lower limbs, J. Neurol. Neurosurg. Psychiat., 23 (1960) 222227. 16 Hollinshead, W., Anatomy for Surgeons: The Head and Neck, Vol. 1, Harper and Row, New York, 1968. 17 Hosokawa, H., Proprioceptive innervation of striated muscles in the territory of cranial nerves, Tex. Rep. Biol. Med., 19 (1961) 405-464. 18 Huntsman, L., Nichols, G., Verdugo, P. and Pollack, G., A new instrument for mechanical testing of tissues, Proc. Ann. Conf. Eng. Med. Biol., 24 (1971) 50. 19 Kadanoff, D., Die sensiblen narvendigungen in der mimshen muskulatur des menschen, Z. mikr.-anat. Forsch., 62 (1956) 1-15. 20 Larson, C., Folkins, J., McClean, M. and Muller, E., Sensitivity of the human perioral reflex to parameters of mechanical stimulation, Brain Research, 146 (1978) 159-164. 21 Limanskii, Y., Pilyavskii, A. and Gura, E., Postsynaptic potentials of facial nucleus motoneurons evoked by afferent and corticofugal impulses, Neurophysiology, 5 (1973) 296-303. 22 McClean, M. and Cooker, H., Preliminary observations on the response of perioral single motor units to mechanical stimulation, J. acoust. Soc. Amer, 67 (1980) 271-275. 23 Milner-Brown, H., Stein, R. and Yemm, R., The contractile properties of human motor units during voluntary isometric contractions, J. Physiol. (Lond.), 228 (1973) 285-306. 24 Penders, C. and Delwaide, P., Physiologic approach to the human blink reflex. In J. Desmedt (Ed.), New Developments in Electromyography and Clinical Neurophysiology, VoL 3, Karger, Basel, 1973, pp. 649-657. 25 Shahani, B. and Young, R., Studies of the normal human silent period. In J. Desmedt (Ed.), New Develepments in Electromyography and Clinical Neurophysiology, VoL 3, Karger, Basel, 1973, pp. 641-648. 26 Smith, A., Zimrnermann, G. and Abbas, P., Recruitment of motor units in speech production, J. Speech Res., 24 (1981) 567-575. 27 Stephens, J., Usherwood, T. and Garnett, R., Technique for studying synaptic connections of single motoneurons in man, Nature (Lond.), 263 (1976) 343-344.
1 Ashby, P. and Labelle, K., The effects of extensor and flexor group I afferent volleys on the excitability of individual soleus motoneurons in man, J. Neurol. Neurosurg. Psychiat., 40 (1977) 910-919. 2 Bak, M. and Sehmidt, E., An improved time-amplitude window discriminator, IEEE Trans. Biomed. Eng., 24 (1977) 486--489. 3 Bratzlavsky, M., Human bralnstem reflexes. In M. Shahani (Ed.), The Motor System: Neurophysiology and Muscle Mechanisms, Elsevier, Amsterdam, 1976, pp. 135-154. 4 Buller, N., Garnett, R. and Stephens, J., The reflex responses of single motor units in human hand muscles following muscle afferent stimulation, J. Physiol. (Lond.), 303 (1980) 337-340. 5 Caccia, M., McComas, A., Upton, A. and Blogg, T., Cutaneous reflexes in small muscles of the hand, J. NeuroL Neurosurg. Psychiat., 36 (1973) 960-972. 6 Darian-Smith, I., Neural mechanisms of facial sensation, Int. Rev. Neurobiol., 9 (1973) 301-395. 7 Dubner, R., Sessle, B. and Storey, A., The Neural Basis of Oral and Facial Function, Plenum Press, New York 1978, 483 pp. 8 Ekbom, K., Jernelius, M. and Kugelberg, E., Perioral reflexes, Neurology, 2 (1952) 103-111. 9 Ellaway, P., Cumulative sum technique and its application to the analysis of peristimulus time histograms, Electroenceph. Clin. Neurophysiol., 45 (1978) 302-304. 10 Erzurumlu, R. and Killackey, H., Efferent connections of the brainstem trigaminal complex with the facial nucleus of the rat, J. comp. NeuroL, 188 (1979) 75-86. 11 Gandiglio, C. and Fra, L., Further observations on facial reflexes, J. neuroL Sci., 5 (1967) 173-185. 12 Garnett, R. and Stephens, J., The reflex response of single motor units in human first dorsal interosseous muscle following cutaneous afferent stimulation, J. Physiol. (Lond.), 303 (1980) 351-364. 13 Gobel, S., Some considerations of synaptic organization in the trigeminal sensory nuclei of the adult cat. In B. Sessle and A. Hannam (Eds.), Mastication and Swallowing: Biological and Clinical Correlates, University Toronto Press, Toronto, 1976, pp. 351-364. 14 Hagbarth, K., Excitatory and inhibitory skin areas for flexor and extensor motoneurones, Acta physiol.
of
Washington
Human
Subjects
Review
Committee.