- Email: [email protected]
Characteristics of the muscle spindle endings of the masticatory muscles in the rabbit under halothane anesthesia
Brain Research 833 Ž1999. 1–9 www.elsevier.comrlocaterbres
Research report
Characteristics of the muscle spindle endings of the masticatory muscles in the rabbit under halothane anesthesia Takafumi Kato, Yuji Masuda, Osamu Hidaka, Akira Komuro, Tomio Inoue, Toshifumi Morimoto
)
Department of Oral Physiology, Faculty of Dentistry, Osaka UniÕersity, 1-8, Yamadaoka, Suita, Osaka, 565 Japan Accepted 2 March 1999
Abstract To explore the response characteristics of muscle spindle units in the masticatory muscles in the rabbit, the responses of muscle spindle units were recorded from the mesencephalic trigeminal nucleus ŽMesV. under halothane anesthesia during ramp-and-hold stretches. Three firing patterns, initial burst ŽIB. at the onset of the dynamic phase, negative adaptation ŽNA. at the end of the dynamic phase and firing during the release ŽFDR. phase, were observed during muscle stretch. IB was present at higher stretch velocities, FDR at lower stretch velocities. The velocity at which an IB or FDR was present was different from unit to unit. Because, within the range of the velocities of stretch tested, units with NA always showed NA and units without NA never did, all recorded units were divided into two groups on the basis of the existence of NA ŽNAŽq. or NAŽy. units.. Response characteristics of the two groups were then compared. NAŽq. units showed an IB more frequently and FDR less frequently than NAŽy. units. NAŽq. units had significantly higher dynamic responsiveness and discharge variability than NAŽy. units. The conduction velocity of the afferents of NAŽq. units was higher than that of NAŽy. units. However, distributions of these measurements were not bimodal. These results suggest that NA is the useful criteria to classify the muscle spindle endings in the masticatory muscles in the rabbit under halothane anesthesia. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Muscle spindle; Rabbit; Trigeminal system; Firing pattern; Classification; Halothane anesthesia
1. Introduction Previous studies have revealed that sensory inputs from the muscle spindle receptor and the periodontal receptor were important in regulating masticatory movement and masticatory force w13,20,26x. Our preliminary experiment in which unitary activities were recorded from the mesencephalic trigeminal nucleus ŽMesV. during cortically-induced rhythmic jaw movements ŽCRJMs. in the rabbit, revealed that muscle spindle units discharged not only during the jaw-opening phase but also during the jaw-closing phase w27x. Because of the lack of classification of the jaw-muscle spindle endings in the rabbit, it remains unclear which type of input from the two types of sensory endings, primary or secondary, is more dominant for control of masticatory force. Classification of muscle spindle afferents of jaw-closing muscles has been employed mainly in cats, not in rabbits. Classification of the jaw-muscle
) Corresponding author. [email protected]
Fax:
q 81-6-879-2885;
E-m ail:
spindle endings in the rabbit is thus necessary to clarify the regulatory mechanisms of mastication. Conventionally, the classification of muscle spindle afferents in the hind limb has been based on the conduction velocity of afferents, because the distribution of conduction velocities is clearly bimodal w23x. In the trigeminal system, because of its anatomical disadvantages, conduction velocity does not have a bimodal distribution and is not appropriate for use in classification w1x. Therefore, the classification of the jaw-muscle spindle endings has been based on their sensitivity to vibration or the effects of succinylcholine w7,16,39x. These experiments were performed under deep anesthesia or pharmacological de-efferentation. Identification of the muscle spindle endings under such conditions cannot be directly related to the evaluation of their functional roles during CRJMs w27x. Previous studies revealed that response patterns to muscle stretch could be a useful criteria for classification w4,11,35,36x. However, exploration of the firing patterns in response to muscle stretch has not been undertaken with respect to the jaw muscles. The aim of our study was to classify muscle spindle afferents of the jaw-closing mus-
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 1 3 5 0 - 5
2
T. Kato et al.r Brain Research 833 (1999) 1–9
cles on the basis of their responses to muscle stretch in the rabbit under halothane anesthesia.
2. Materials and methods 2.1. Surgical procedure Seventy-five male rabbits weighing 2.5–3.5 kg were used. Surgical procedures were approved by the Osaka University Faculty of Dentistry Intramural Animal Care and Use Committee. The animals were initially anesthetized with intravenous administration of ketamine Ž16 mgrkg. and sodium thiamylal Ž20 mgrkg. via the ear vein and the trachea was cannulated to supplement a mixture of halothane and oxygen Ž1.5–2%, 2 lrmin.. During the surgery, anesthesia was maintained at such a level that no jaw-opening reflex resulted from pinching the skin. All sites of incision were infiltrated with 2% xylocaine hydrochloride. The rectal temperature was maintained between 36–388C with a heating pad. An electrocardiogram was continuously monitored with heart rates ranging from between 180 to 210 timesrmin. Small screws were implanted in the mentum to hold a phototransistor array with dental acrylic resin. This phototransistor was used to trace jaw movements by means of an optoelectronic recording device ŽC2399, Hamamatsu Photonics, Hamamatsu, Japan.. The attachments to connect the dual mode mechanical controller ŽDPS-280, Dia Medical System, Tokyo, Japan. for jaw opening were fixed on the same screws with dental acrylic resin. The skin flap of the left cheek was reflected, and the surface of the masseter muscle and the medial pterygoid muscle was covered with liquid paraffin and the skin flap. These muscles were exposed only when gentle probing to identify the muscle origin of the afferents was performed. For EMG recording, pairs of enamel-coated copper wires Ž100 mm diameter, 10 mm spacing, 1.5 mm tip bared. were inserted into the anterior part of the left deep masseter muscle. Three screws were inserted into the dorsal aspect of the skull to mount a stereotaxic frame. For recording muscle spindle units from MesV, the dura was exposed at the occipital area centered at 14 mm posterior to bregma and 2 mm left of the midline. In 18 rabbits, conduction velocity was measured. A branch of masseter nerve was stimulated by the bipolar electrode set at the position where the masseter nerve fibers run under the zygomatic arch, a bone which was removed as carefully as possible. The electrode was covered with Vaseline and a 0.05 ms 70–120 mA square pulse was delivered. 2.2. Unit recording A glass-coated elsiloy microelectrode with impedance 1–3 MV at 1 kHz was inserted vertically through the left cerebral cortex towards the MesV. During the recording, anesthesia Žhalothane, 1.5%, 2 lrmin. was maintained at
such a level that neither a corneal reflex nor a withdrawal reflex from pinching occurred. In addition, neither spontaneous muscle activity nor a jaw-closing reflex resulting from jaw opening was observed. This was done to ensure that the masticatory muscles would be relaxed avoiding loading muscle spindle afferents by concomitant extrafusal muscle contraction and thereby disturbing jaw movement. 2.3. Response to muscle stretch Mechanical stimulation was applied by means of a dual mode mechanical controller attached to the connector in the mentum. The jaw movement applied in the present study was performed in the ramp-and-hold stretch manner. The initial jaw position, at which upper and lower teeth did not contact, was 2.58 from the fully closed position. The center of rotation of the lower jaw was referred to in the report from Weijs et al. w43x. The holding phase lasted 1.5 s and the cycle was repeated every 6 s. Stretch and release speed were the same within each ramp stretch and were changed as required. For each unit, three types of parameters of the response to muscle stretch were defined as follows: Ž1. initial burst ŽIB. was defined as the high instantaneous frequency of firing at the start of the muscle stretch; Ž2. negative adaptation ŽNA. was defined as the abrupt fall in frequency at the beginning of the hold phase, in case that the instantaneous frequency of the first impulse at the beginning of hold phase was lower than that of following impulses during subsequent 100 ms; and Ž3. the firing during release ŽFDR. phase. The dynamic and static responsiveness to muscle stretches were calculated as follows. Two basic measurements were calculated from five muscle stretches and the averages were used in the analysis ŽFig. 1.. The dynamic index ŽDI. was defined as the difference between the mean frequency obtained from the 0.05 s period preceding the end of the dynamic phase and the frequency obtained from the 0.4 s to 0.6 s period after the end of the dynamic phase. The static difference ŽSD. was defined as the difference between the mean frequency obtained from 0.4 s to 0.6 s period after the end of the dynamic phase and the 0.5 s period preceding the ramp stretch onset. Dynamic responsiveness was evaluated by both the DI and the velocity sensitivity ŽVS.. The VS was determined from the slope of the regression line obtained from the linear relationships between the five stretch rates Ž6, 12, 15, 20 and 308rs, with 38 amplitude. and the DIs at each stretch rate. The static responsiveness was evaluated by the position sensitivity ŽPS.. The SD at five stretch amplitudes Ž1, 1.8, 3, 4.8 and 68, at 128rs. was measured and the slope of the regression line between the SDs and the stretch amplitudes was defined as the PS. Although the head of the animal was fixed as firmly as possible, the unitary recordings sometimes failed when the stretches with the fast rates Ž208rs and more. or the large amplitudes Ž4.88 and more. were applied. That is the reason why
T. Kato et al.r Brain Research 833 (1999) 1–9
3
Fig. 1. Unit response to jaw opening at 128rs with 38 amplitude. Top trace: unit discharges. Middle trace: instantaneous frequency. Bottom trace: vertical jaw movement. Upward arrow indicates the direction of jaw opening. DI: dynamic index. SD: static difference.
the number of units stated in the results varies. Discharge variability was defined as the coefficient of variation ŽŽS.D.rmean interspike intervals.100. during the last 500 ms in the hold phase of the stretch with a 38 amplitude.
discrimination and calculation were performed using Spike2 software ŽCambridge Electronic Design..
2.4. Conduction Õelocity of the spindle afferents
To compare two groups, the null hypothesis that each variable had a normal distribution was initially tested by a chi-square test. Then, the null hypothesis that the variance had homogeneity was tested by an F-test. If the two null hypotheses were not rejected, the difference between two groups was tested by a two sample t-test. When either or none of the above two null hypotheses were not rejected, the difference was tested by a Mann–Whitney test. In a comparison among groups, the difference was tested by a Kruskal–Wallis test. Results are presented as mean " S.D. In some cases, to test the assumption that the frequency distribution of quantitative data consists of two subpopulations, data were analyzed by the statistical method of Hald w12x. Mean values and S.D.s for the two possible subpopulations were defined and presented as mean " S.D. This method was used the previous reports w34,39x.
The conduction velocity was calculated from the latency of the evoked action potential and the conduction distance. The distances between the stimulating site to the MesV were measured after post-fixation. 2.5. Histology At the end of the experiments, the animals were killed by an overdose of pentobarbital sodium administered intravenously. Recording sites were marked by passing a negative current of 30 mA through the recording electrode for 10 s. The animals were perfused with saline followed by Formalin Ž10%.. After post-fixation, the brain was removed from the skull and frozen sections were cut in the frontal plane and stained with Cresyl violet. Recording sites were identified within MesV under light microscope.
2.7. Statistical analysis
3. Results 2.6. Data analysis 3.1. Identification of the spindle afferents The unitary discharge, the jaw movements and EMG activity were simultaneously recorded on an 8-channel DAT data recorder ŽPC-208A, Sony-Magnescale, Tokyo, Japan.. The data was replayed and fed into a personal computer ŽDell Optiplex XMT 5133, Dell Computer, Round Rock, TX, USA. with CED 1401 plus interface ŽCambridge Electronic Design, Cambridge, UK.. Spike
Responses of 168 units were recorded from the MesV. The location of the units was confirmed histologically. These units responded to the jaw opening, but did not respond to pressure applied to the teeth, gingiva or vibrissae. Furthermore, the first-order neurons of golgi tendon organs are located in Gasserian ganglia rather than in the
4
T. Kato et al.r Brain Research 833 (1999) 1–9
MesV w21x. Accordingly, all units recorded here were muscle spindle units. The muscle origins of 156 units were identified by gentle muscle probing. Ninety-seven units innervated in the masseter muscle, 19 in the medial pterygoid muscle and five in the temporal muscle. Another 35 units did not respond to muscle probing but did respond to pushing on the eyeball. Because the posterolateral bone of the orbital fossa does not exist in the rabbit, it is assumed that by pushing on the eye-ball some soft tissue around the eye-ball indirectly pushed the deep parts of the masseter muscles or the pterygoid muscles and the spindles were stretched. Most of the units innervating the masseter muscle responded to medial jaw shift and those in the medial pterygoid muscle responded to lateral jaw shift. There were no significant difference in all measurements calculated in the present study among the muscles of origin Ž p ) 0.05, Kruskal–Wallis test.. 3.2. Firing pattern The response to ramp-and-hold stretches was investigated with respect to three qualitative parameters, i.e., IB, NA, and FDR. These parameters are thought to be useful in discriminating between primary and secondary muscle spindle endings w4,35,36x. Among the three parameters, IB and FDR were more related to the rate of the change in muscle length than NA. Fig. 2 illustrates this relationship.
The unit in Fig. 2A did not show an IB in response to stretch at 128rs while an IB was present at 308rs. On the other hand, the unit in Fig. 2B showed FDR at the velocity of release in 128rs, but not FDR at 308rs. Within the range of the velocities in the present study Ž6, 12, 15, 20 and 308rs., the threshold of the velocity of stretch or release at which an IB appeared or the FDR disappeared varied among the units. However, units which showed NA in their response to stretch at 128rs always showed NA at faster velocities than 128rs, and vice versa ŽFig. 2.. In the case when the much faster velocity of stretch and release Ž608rs. was applied to 62 units, units without NA at 308rs never showed NA at 608rs. As a result, the decision whether each unit has NA or not was made most confidently among three parameters, although it is known that the characteristics of these three parameters are dependent on the rate of change of muscle length w14,19,23,33x. Furthermore, there was evidence that NA could be a reliable criterion w4,14,15x. Therefore, all recorded units were divided into two groups on the basis of the existence of NA at the velocity of stretch of 128rs. Out of 168 units, 61 units showed NA ŽNAŽq. units. and 107 units did not ŽNAŽy. units. at the velocity of 128rs. Table 1 shows the number of units with an IB and FDR at two velocities Ž12 and 308rs. and compares NAŽq. and NAŽy. units. At both velocities, NAŽq. units were likely to show more IB and less FDR than NAŽy. units Žin all cases: p - 0.001, chi-square test for independence..
Fig. 2. Response patterns of two units at two different velocities of stretch Ž128rs and 308rs.. A,B: Response of an NAŽq. unit. Although NA existed at both velocities of stretch, an IB was not present at 128rs ŽA. but was at 308rs ŽB.. C,D: Response of an NAŽy. unit. FDR was observed at 128rs ŽC. and disappeared at 308rs ŽD.. Within the range of the velocities of stretch examined in the present study, NA never appeared in this type of unit. Downward arrow: IB, upward open arrow: NA, asterisk: FDR.
T. Kato et al.r Brain Research 833 (1999) 1–9 Table 1 Contingency table of firing patterns at 128rs and 308rs 128rs
IB FDR
Žq. Žy. Žq. Žy.
308rs
NAŽq. Ž ns61.
NAŽy. Ž ns107.
NAŽq. Ž ns 41.
NAŽy. Ž ns65.
13 48 8 53
3 104 69 38
28 13 0 41
7 58 22 43
IB was more frequent in the response of NAŽq. units than NAŽy. units and a higher proportion of NAŽq. units showed IB at both velocities. In contrast, FDR was more frequent in the response of NAŽq. units than NAŽy. units at both velocities.
3.3. QuantitatiÕe analysis The distributions of the DIs for 168 units at 128rs and for 106 units at 308rs are shown in Fig. 3A and B. The
5
DIs were 49.5 " 24.3 impulsesrs for 128rs and 70.8 " 33.5 impulsesrs for 308rs. The DIs at 128rs for 61 NAŽq. units Ž69.2 " 20.4 impulsesrs. were significantly higher Ž p - 0.001; two sample t-test. than those for 107 NAŽy. units Ž38.3 " 18.6 impulsesrs.. A significant difference in DIs at 308rs between the two groups was also observed ŽNAŽq. units: 101.0 " 27.9 impulsesrs, NAŽy. units: 51.8 " 20.2 impulsesrs, p - 0.001, Mann–Whitney U-test.. The distributions of DIs at the two velocities of stretch were not clearly bimodal and there was a large overlap between the two groups. However, the results from Hald’s method revealed that at each velocity of stretch the distribution of DI consisted of two subpopulations. In Fig. 3C and D, two curves of two subpolulations defined by Hald’s method ŽC: 74.1 " 21.1 impulsesrs and 36.1 " 16.7 impulsesrs, D: 115.4 " 34.1 impulsesrs and 51.6 " 18.8 impulsesrs. are shown. In both C and D, it is apparent that
Fig. 3. Histograms of the DI calculated from the responses of 168 units at 128rs ŽA. and from 106 of those units at 308rs ŽB.. Filled bars indicate NAŽq. units and open bars indicate NAŽy. units. At both velocities of stretch, NAŽq. units had a significantly higher DI than those of NAŽy. units ŽA: p - 0.001, two sample t-test, B: p - 0.001, Mann–Whitney U-test.. In C and D, histograms of total distributions of the DI corresponding to A and B, respectively, are shown. Two curves in each histogram indicate frequency distributions of two subpopulations calculated by the method of Hald.
6
T. Kato et al.r Brain Research 833 (1999) 1–9
The VS was determined to evaluate the sensitivity of 106 units to the change of the five different velocities of stretch. VS was distributed mainly from 0.07 impulsesrdegree to 2.63 impulsesrdegree with a maximum of 4.19 impulsesrdegree ŽFig. 4.. The VS of 41 NAŽq. units was 1.85 " 0.55 impulsesrdegree and that of 65 NAŽy. units was 0.83 " 0.41 impulsesrdegree. The VS of the NAŽq. group was significantly larger than that of the NAŽy. group Ž p - 0.001, Mann–Whitney U-test.. The result obtain from Hald’s method is shown in Fig. 4B. Although the frequency distribution of NAŽq. units Žfilled bars. was not normally distributed because of one extreme data point, a correlation between the two curves Ž2.23 " 0.74 impulsesrdegree and 0.81 " 0.5 impulsesrdegree. and the two actual distributions was observed. The VS in the present study was lower than in the masseter muscle of the cat with the mean 30 impulsesrdegree w16x. PS was determined to evaluate the sensitivity of units to the change of the amplitudes of the jaw opening. The
Fig. 4. A: Histogram of the VSs calculated from 106 units ŽA.. Filled bars indicate NAŽq. units Ž ns 41. and open bars indicate NAŽy. units Ž ns65.. VS of NAŽq. units was significantly higher than NAŽy. units Ž p- 0.001, Mann–Whitney U-test.. In B, histograms of total distributions of the VS are shown. Two curves indicate frequency distributions of two subpopulations calculated by the method of Hald.
there is much correspondence with the actual frequency distributions of two groups defined as NAŽq. units Žfilled bars. and NAŽy. units Žopen bars..
Fig. 5. Histogram of the PSs calculated from 59 units. Filled bars indicate NAŽq. units Ž ns 24. and open bars indicate NAŽy. units Ž ns 35.. There was no significant difference between the two types of unit groups.
Fig. 6. A: Histogram of the discharge variability calculated from 168 units. Filled bars indicate NAŽq. units Ž ns61. and open bars indicate NAŽy. units Ž ns107.. The discharge variability of NAŽq. units was significantly higher than NAŽy. units Ž p- 0.001, Mann–Whitney Utest.. In B, histograms of total distributions of the discharge variability are shown. Two curves indicate frequency distributions of two subpopulations calculated by the method of Hald. Bin width of the histogram in B was defined as 3 to increase the number of data points for this calculation.
T. Kato et al.r Brain Research 833 (1999) 1–9
distribution of PS obtained from 59 units is shown in Fig. 5. The mean PS calculated from these units was 16.2 " 5.6 impulsesrs per degree, 17.2 " 6.3 impulsesrs per degree for 24 NAŽq. units and 15.5 " 5.2 impulsesrs per degree for 35 NAŽy. units. There was no significant difference between the two groups. PS in the present study was much higher than that found in cats where a mean of 7.55 impulsesrs per degree was found w16x. Discharge variability for 168 units was determined to compare the regularity of firing during the hold phase between the two groups. The mean discharge variability obtained from 61 NAŽq. units was 23.0 " 9.9 and that of 107 NAŽy. units was 8.2 " 4.3. NAŽy. units discharged significantly Ž p - 0.001, Mann–Whitney U-test. more regularly than NAŽq. units ŽFig. 6.. Two subpopulations determined by the Hald’s method Ž27.0 " 11.5 and 7.7 " 3.8. had a good correlation with the actual two populations of NAŽq. and NAŽy. units ŽFig. 6B.. Some NAŽq. units fired with a small number of spikes during the last 500 ms of the hold phase, which made their discharge variability large. However, even when this source of variability is taken into account, the discharge variability was much larger in the present study than in results reported elsewhere w7,24,31x. Residual fusimotor activity may also be attributed to relatively large discharge variability and PS. Thirty-seven units that followed a repetitive electrical stimulation to the masseter nerve at 333 Hz or above were identified as masseter spindle units. These units responded to muscle stretch and were located in the MesV. Conduction velocities of 37 units ranged from 17.9 mrs to 54.5 mrs with a mean value of 40.0 " 9.2 mrs. As shown in Fig. 7, conduction velocities of the two unit groups heavily overlapped; however, the conduction velocity of 17 NAŽq. units Ž45.3 " 5.1 mrs. was significantly faster than that of 20 NAŽy. units Ž35.5 " 9.5 mrs. Ž p - 0.001, Mann– Whitney U-test..
Fig. 7. Histogram of the conduction velocities calculated from 37 units innervating the masseter muscle.
7
4. Discussion The results revealed that NAŽq. and NAŽy. units had different response characteristics. NAŽq. units showed an IB at the start of stretch more frequently and discharged during the release phase less frequently than NAŽy. units. Dynamic responsiveness, assessed by measuring DI, VS, and discharge variability were higher in NAŽq. units than in NAŽy. units. In the hind limb muscle of the cat, physiological classification of muscle spindle afferents has been employed on the basis of conduction velocity because the distribution of conduction velocities is clearly bimodal. The primary endings are formed by group Ia afferents with larger axon diameters and higher conduction velocities, and the secondary endings by group II afferents with smaller axon diameters and lower conduction velocities w23x. The distribution of conduction velocities of the muscle spindle afferents in the hind limb of the rabbit was also bimodal w10,28x. On the other hand, in masticatory muscles of the cat, the distributions of both axon diameters and conduction velocities of muscle spindle afferents are not clearly bimodal w2,16,25x. In the present study, the distribution of conduction velocities of jaw-muscle spindle afferents in rabbits was not bimodal, either, although there was a significant difference between NAŽq. and NAŽy. units. Therefore, like in cats, it is impossible to classify the muscle spindle afferents in the masticatory muscle of the rabbit on the basis of the conduction velocity. Previous studies revealed that the response to the rampand-hold stretch was different between primary and secondary endings w4,8,23,34,35x. The primary ending shows a high frequency of discharge at the start of the dynamic phase ŽIB., an abrupt fall of discharge at the start of the hold phase ŽNA., and is likely to fall silent during the release phase. Classification based on these response patterns has been reported in some studies w4,35x, but not in the jaw muscles. Cheney and Preston w4x reported that, in the soleus muscle of the baboon, the classification based on NA was consistent with that of the conduction velocity of the afferent. Schafer ¨ w34x reported that in the cat hind limb the afferents with conduction velocity ranging from 80 to 105 mrs showed NA in response to ramp stretch. Hunt and Ottoson w14x recorded the afferent response to a ramp-and-hold stretch from individual axons with primary and secondary endings in an isolated muscle spindle preparation. They found that the rapid fall of the receptor potential at the end of the dynamic phase was evident at a high velocity of stretch only in the primary endings. The rapid fall of the receptor potential could cause the abrupt fall in discharge, NA w15x. These studies suggest that NA may be a useful criteria to classify muscle spindle endings. Since the appearance or amplitude of the IB is dependent on the stiffness of intrafusal muscle fibers w30x, Scott w35x tested the classification by using the IB or combined criteria with the IB, NA and FDR in the cat. They condi-
8
T. Kato et al.r Brain Research 833 (1999) 1–9
tioned the parent muscle to enhance the amplitude of the IB by electrical stimulation to the efferents. They reported that this conditioning procedure improved the discrimination between the two types of afferents. In the present study, conditioning procedures were not performed because it was difficult to stimulate the efferents innervating all the masticatory muscles. This might have caused fewer units with IB in their responses to stretch than in the report by Scott w35x. The distributions of quantitative measurements were not bimodal and there was a large overlap between the two groups. This result seems to be common in the jaw muscles w4,16,39x, and is attributed to the histological differences between the jaw and limb muscles or to the difference in the experimental condition between the present and previous studies. Compared with the cat’s hind limb muscle spindle, there was a high proportion of intermediate spindles in the jaw-muscle spindle in the cat w16,39x. A high proportion of intermediate muscle spindles is related to the existence of complexes of spindles and tandem spindles w21x, and to the various innervation patterns to those spindles w3,32x. In the jaw muscles of the rabbits, however, spindles exist more simply than in the cat w32x. Spindles usually occurred singly, rarely as partially fused pairs, with one bag1, one bag2, and two to seven chain fibers. This histological study may result in less overlap between measurements of the two groups in the rabbit than in the cat. Other potential factors contributing to the large overlap may be various innervation patterns like branching of jaw-muscle spindle afferents outside the capsule w17x and separate innervation to bag1 and bag2 fibers w3x. In addition, the experimental conditions used in this study vs. previous studies should be considered. In rabbits, unitary recordings often failed when the stretch with large amplitude was applied because fixation of the head was less rigid than in cats. Therefore, in the present study, the muscle was stretched at an intermediate amplitude. It was reported that some of the primary units did not show NA in other muscles, for example, the baboon w19x, human w11x, and cat neck muscles w31x. In the study by Edin and Vallbo w11x, primary afferents lacked NA more frequently in humans than in the baboon. They assumed that their result was owing to the fact that the muscles were at an intermediate length during ramp stretch. To help relate the results from this study to the experiments reported previously w13,27x, we chose halothane anesthesia. Pages and Proske w29x suggested that halothane might have a direct action on the receptor, because the dynamic response of primary endings was enhanced by halothane inhalation under deep anesthesia with de-efferentation. It was shown that the effects of electrical stimulation in the different central structures to the responses of muscle spindle endings were modulated by increasing the depth of halothane anaesthesia w40–42x. They did not discuss specific fusimotor activity caused by
halothane anaesthesia. From their results, it is assumed that halothane anesthesia may have had an effect on fusimotor activity by modulating the central nervous system. Under the de-efferented condition, discharge variability of muscle spindle endings in different muscles was generally very low w7,24,31x. It was increased strongly by static fusimotor activity w5,24x. Since discharge variability in this study was in a higher range than those studies with de-efferented or fusimotor-suppressed conditions w7,24,31x, residual static fusimotor activity was likely present in our preparation. It remains unclear how the interaction of central static effects with peripheral dynamic effects produce changes in fusimotor activity and thereby affect the sensitivity of muscle spindle endings. However, since these fusimotor activities modulate the response characteristics of muscle spindle afferents as described previously w5,23x, they could affect the value of quantitative data and then contribute to the large overlap between the two groups. Despite the presence of some confounding factors for classification as discussed above, the present results clearly revealed that the two types of units, divided with respect to the existence of NA, had different response characteristics to muscle stretch. It could be inferred that the units with NA may be regarded as primary endings and units without NA as secondary endings. In our previous report, it was observed that there might be a difference in the response of muscle spindle endings to the different hardness of the test materials used during CRJMs under halothane anesthesia w27x. During mastication in awake animals or during reflexively-evoked jaw movements, different behaviors between dynamic and non-dynamic types of units were observed w6,22,37,38x. Morphological studies revealed that the dynamic–sensitive muscle spindle afferents in the cat and rat jaw muscles project differently to the trigeminal motor nucleus and surrounding areas from the non-dynamic muscle spindle afferents Žrat w9x; cat w18x.. Sensory inputs from jaw-muscle spindles are thought to be importantly involved in regulating masticatory force w13x. However, behavior of the jaw-muscle spindle endings during CRJMs has not been clearly demonstrated and the role of proprioceptive inputs to control masticatory force during CRJMs remains uncertain. In consideration of the studies mentioned above, it is assumed that the two types of unit groups observed in the present study play different roles in the regulation of masticatory force. References w1x K. Appenteng, Jaw muscle spindle and their central connections, in: A. Taylor ŽEd.., Neurophysiology of the Jaws and Teeth, Macmillan, London, 1990, pp. 247–255. w2x K. Appenteng, T. Morimoto, A. Taylor, Fusimotor activity in masseter nerve of the cat during reflex movements, J. Physiol. 305 Ž1980. 415–431. w3x R.W. Banks, D. Barker, H.H. Saed, M.J. Stacey, Innervation of muscle spindles in rat deep masseter, J. Physiol. 406 Ž1988. 161P. w4x P.D. Cheney, J.B. Preston, Classification and response character-
T. Kato et al.r Brain Research 833 (1999) 1–9
w5x w6x
w7x
w8x
w9x
w10x
w11x w12x w13x
w14x
w15x
w16x
w17x
w18x
w19x w20x
w21x
w22x
w23x
istics of muscle spindle afferents in the primate, J. Neurophysiol. 39 Ž1976. 1–8. P.D. Cheney, J.B. Preston, Classification of fusimotor fibers in the primate, J. Neurophysiol. 39 Ž1976. 9–19. F.W.J. Cody, L.M. Harrison, A. Taylor, Analysis of the activity of jaw-closing muscles during normal movements in the cat, J. Physiol. 252 Ž1975. 565–592. F.W.J. Cody, R.W.H. Lee, A.A. Taylor, Functional analysis of the components of the mesencephalic nucleus of the fifth nerve in the cat, J. Physiol. 226 Ž1972. 249–261. S. Cooper, The responses of the primary and secondary endings of muscle spindles with intact motor innervation during applied stretch, J. Exp. Physiol. 46 Ž1961. 383–398. D. Dessem, R. Donga, P. Luo, Primary- and secondary-like jawmuscle spindle afferents have characteristic topographic distribution, J. Neurophysiol. 77 Ž1997. 2925–2944. K. Diete-Spiff, The discharge pattern of muscle spindles of the rabbit on activation of intrafusal muscle fibres, J. Physiol. 159 Ž1961. 282–296. ˚ Vallbo, Dynamic response of human muscle spindle B.B. Edin, A.B. afferents to stretch, J. Neurophysiol. 63 Ž1990. 1297–1306. A. Hald, Statistical Theory with Engineering Applications, Wiley, New York, 1952. O. Hidaka, T. Morimoto, Y. Masuda, T. Kato, R. Matsuo, T. Inoue, M. Kobayashi, K. Takada, Regulation of masticatory force during cortically induced rhythmic jaw movements in the anesthetized rabbit, J. Neurophysiol. 77 Ž1997. 3168–3179. C.C. Hunt, D. Ottoson, Impulse acitivity and receptor potential of primary and secondary endings of isolated mammalian muscle spindles, J. Physiol. 252 Ž1975. 259–281. C.C. Hunt, R.S. Wilkinson, Y. Fukami, Ionic basis of the receptor potential in primary endings of mammalian muscle spindles, J. Gen. Physiol. 71 Ž1978. 683–698. H. Inoue, T. Morimoto, Y. Kawamura, Response characteristics and classification of muscle spindles of the masseter muscle in the cat, Exp. Neurol. 74 Ž1981. 548–560. T. Kato, Y. Kawamura, T. Morimoto, Branching of muscle spindle afferents of jaw closing muscles in the cat, J. Physiol. 323 Ž1982. 483–495. H. Kishimoto, Y.C. Bae, A. Yoshida, M. Moritani, M. Takemura, S. Nakagawa, Y. Nagase, T. Wada, B.J. Sessle, Y. Shigenaga, Central distribution of synaptic contacts of primary and secondary jaw muscle spindle afferents in the trigeminal motor nucleus of the cat, J. Comp. Neurol. 391 Ž1998. 50–63. T.H. Koeze, Muscle spindle afferent studies in the baboon, J. Physiol. 229 Ž1973. 297–317. G. Lavigne, J.S. Kim, C. Valiquette, J.P. Lund, Evidence that periodontal pressoreceptors provide positive feedback to jaw closing muscles during mastication, J. Neurophysiol. 58 Ž1987. 342–358. J.P. Lund, F.J.R. Richmond, C. Touloumis, Y. Patry, Y. Lamarre, The distribution of Golgi tendon organs and muscle spindles in masseter and temporalis muscles of the cat, Neuroscience 3 Ž1978. 259–270. Y. Masuda, T. Morimoto, O. Hidaka, T. Kato, R. Matsuo, T. Inoue, M. Kobayashi, A. Taylor, Modulation of jaw muscle spindle discharge during mastication in the rabbit, J. Neurophysiol. 77 Ž1997. 2227–2231. P.B.C. Matthews, Mammalian Muscle Receptors and Their Central Actions, Edward Arnold, London, 1972.
9
w24x P.B.C. Matthews, R.B. Stein, The regularity of primary and secondary muscle spindle afferent discharges, J. Physiol. 202 Ž1969. 59–82. w25x T. Morimoto, H. Inoue, Y. Kawamura, Diameter spectra of sensory and motor fibers in nerves to jaw-closing and jaw-opening muscles in the cat, Jpn. J. Physiol. 33 Ž1982. 171–182. w26x T. Morimoto, T. Inoue, Y. Masuda, T. Nagashima, Sensory components facilitating jaw-closing muscle activities in the rabbit, Exp. Brain Res. 76 Ž1989. 424–440. w27x T. Morimoto, O. Nakamura, K. Ogata, Z.J. Liu, R. Matsuo, T. Inoue, Y. Masuda, O. Saito, J. Mizuno, T. Kato, Autoregulation of masticatory force in the anesthetized rabbit, in: T. Morimoto, T. Matsuya, K. Takada ŽEds.., Brain and Oral Functions. Elsevier, Amsterdam, 1995, pp. 115–124. w28x P.R. Murphy, The patterns of discharge of static and dynamic g motoneurons in the decerebrate rabbit, J. Physiol. 333 Ž1982. 29–37. w29x B. Pages, U. Proske, Effect of halothane anesthesia on responses from primary endings of muscle spindles in the cat, Exp. Neurol. 28 Ž1970. 393–402. w30x U. Proske, G.J. Stuart, The initial burst of impulses in responses of toad muscle spindles during stretch, J. Physiol. 368 Ž1985. 1–17. w31x F.J.R. Richmond, V.C. Abrahams, Physiological properties of muscle spindles in dorsal neck muscles of the cat, J. Neurophysiol. 42 Ž1979. 604–617. w32x A. Rowlerson, F. Mascarello, D. Barker, H. Saed, Muscle-spindle distribution in relation to the fibre-type composition of masseter in mammals, J. Anat. 161 Ž1988. 37–60. w33x S.S. Schafer, ¨ The acceleration response of a primary muscle-spindle ending to ramp stretch of the extrafusal muscle, Experientia 23 Ž1967. 1027. w34x S.S. Schafer, Regularity in the generation of discharge patterns by ¨ primary and secondary muscle spindle afferents, as recorded under a ramp-and-hold stretch, Exp. Brain Res. 102 Ž1994. 198–209. w35x J.J.A. Scott, Classification of muscle spindle afferents in the peroneus brevis muscle of the cat, Brain Res. 509 Ž1990. 62–70. w36x J.J.A. Scott, Response of Ia afferent axons from muscle spindles lacking a bag1 intrafusal muscle fibre, Brain Res. 543 Ž1991. 97–101. w37x A. Taylor, F.W.J. Cody, Jaw muscle spindle activity in the cat during normal movements of eating and drinking, Brain Res. 71 Ž1974. 523–530. w38x A. Taylor, O. Hidaka, R. Durbaba, P.H. Ellaway, Fusimotor influence on jaw muscle spindle activity during swallowing-related movements in the cat, J. Physiol. 503 Ž1997. 157–167. w39x A. Taylor, R. Durbaba, J.F. Rodgers, The classification of afferents from muscle spindles of the jaw-closing muscles of the cat, J. Physiol. 456 Ž1992. 609–628. w40x J.P. Vedel, Effets des stimulations du noyau caude´ sur l’activite´ des terminaisons fusoriales primaires et secondaires du muscle soleaire, ´ Exp. Brain Res. 22 Ž1975. 113–128. w41x J.P. Vedel, J. Mouillac-Baudevin, Controle ˆ de l’activite´ des fibres fusimotrices dynamiques et statiques par la formation reticulee ´ ´ mesencephalique chez le chat, Exp. Brain Res. 9 Ž1969. 307–324. ´ ´ w42x J.P. Vedel, J. Mouillac-Baudevin, Controle ˆ pyramidal de l’activite´ des fibres fusimotrices dynamiques et statiques chez la chat, Exp. Brain Res. 10 Ž1970. 39–63. w43x W.A. Weijs, J.A.M. Korfage, G.J. Langenbach, The functional significance of the position of the centre of rotation for jaw opening and closing in the rabbit, J. Anat. 162 Ž1989. 133–148.
Research report
Characteristics of the muscle spindle endings of the masticatory muscles in the rabbit under halothane anesthesia Takafumi Kato, Yuji Masuda, Osamu Hidaka, Akira Komuro, Tomio Inoue, Toshifumi Morimoto
)
Department of Oral Physiology, Faculty of Dentistry, Osaka UniÕersity, 1-8, Yamadaoka, Suita, Osaka, 565 Japan Accepted 2 March 1999
Abstract To explore the response characteristics of muscle spindle units in the masticatory muscles in the rabbit, the responses of muscle spindle units were recorded from the mesencephalic trigeminal nucleus ŽMesV. under halothane anesthesia during ramp-and-hold stretches. Three firing patterns, initial burst ŽIB. at the onset of the dynamic phase, negative adaptation ŽNA. at the end of the dynamic phase and firing during the release ŽFDR. phase, were observed during muscle stretch. IB was present at higher stretch velocities, FDR at lower stretch velocities. The velocity at which an IB or FDR was present was different from unit to unit. Because, within the range of the velocities of stretch tested, units with NA always showed NA and units without NA never did, all recorded units were divided into two groups on the basis of the existence of NA ŽNAŽq. or NAŽy. units.. Response characteristics of the two groups were then compared. NAŽq. units showed an IB more frequently and FDR less frequently than NAŽy. units. NAŽq. units had significantly higher dynamic responsiveness and discharge variability than NAŽy. units. The conduction velocity of the afferents of NAŽq. units was higher than that of NAŽy. units. However, distributions of these measurements were not bimodal. These results suggest that NA is the useful criteria to classify the muscle spindle endings in the masticatory muscles in the rabbit under halothane anesthesia. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Muscle spindle; Rabbit; Trigeminal system; Firing pattern; Classification; Halothane anesthesia
1. Introduction Previous studies have revealed that sensory inputs from the muscle spindle receptor and the periodontal receptor were important in regulating masticatory movement and masticatory force w13,20,26x. Our preliminary experiment in which unitary activities were recorded from the mesencephalic trigeminal nucleus ŽMesV. during cortically-induced rhythmic jaw movements ŽCRJMs. in the rabbit, revealed that muscle spindle units discharged not only during the jaw-opening phase but also during the jaw-closing phase w27x. Because of the lack of classification of the jaw-muscle spindle endings in the rabbit, it remains unclear which type of input from the two types of sensory endings, primary or secondary, is more dominant for control of masticatory force. Classification of muscle spindle afferents of jaw-closing muscles has been employed mainly in cats, not in rabbits. Classification of the jaw-muscle
) Corresponding author. [email protected]
Fax:
q 81-6-879-2885;
E-m ail:
spindle endings in the rabbit is thus necessary to clarify the regulatory mechanisms of mastication. Conventionally, the classification of muscle spindle afferents in the hind limb has been based on the conduction velocity of afferents, because the distribution of conduction velocities is clearly bimodal w23x. In the trigeminal system, because of its anatomical disadvantages, conduction velocity does not have a bimodal distribution and is not appropriate for use in classification w1x. Therefore, the classification of the jaw-muscle spindle endings has been based on their sensitivity to vibration or the effects of succinylcholine w7,16,39x. These experiments were performed under deep anesthesia or pharmacological de-efferentation. Identification of the muscle spindle endings under such conditions cannot be directly related to the evaluation of their functional roles during CRJMs w27x. Previous studies revealed that response patterns to muscle stretch could be a useful criteria for classification w4,11,35,36x. However, exploration of the firing patterns in response to muscle stretch has not been undertaken with respect to the jaw muscles. The aim of our study was to classify muscle spindle afferents of the jaw-closing mus-
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 1 3 5 0 - 5
2
T. Kato et al.r Brain Research 833 (1999) 1–9
cles on the basis of their responses to muscle stretch in the rabbit under halothane anesthesia.
2. Materials and methods 2.1. Surgical procedure Seventy-five male rabbits weighing 2.5–3.5 kg were used. Surgical procedures were approved by the Osaka University Faculty of Dentistry Intramural Animal Care and Use Committee. The animals were initially anesthetized with intravenous administration of ketamine Ž16 mgrkg. and sodium thiamylal Ž20 mgrkg. via the ear vein and the trachea was cannulated to supplement a mixture of halothane and oxygen Ž1.5–2%, 2 lrmin.. During the surgery, anesthesia was maintained at such a level that no jaw-opening reflex resulted from pinching the skin. All sites of incision were infiltrated with 2% xylocaine hydrochloride. The rectal temperature was maintained between 36–388C with a heating pad. An electrocardiogram was continuously monitored with heart rates ranging from between 180 to 210 timesrmin. Small screws were implanted in the mentum to hold a phototransistor array with dental acrylic resin. This phototransistor was used to trace jaw movements by means of an optoelectronic recording device ŽC2399, Hamamatsu Photonics, Hamamatsu, Japan.. The attachments to connect the dual mode mechanical controller ŽDPS-280, Dia Medical System, Tokyo, Japan. for jaw opening were fixed on the same screws with dental acrylic resin. The skin flap of the left cheek was reflected, and the surface of the masseter muscle and the medial pterygoid muscle was covered with liquid paraffin and the skin flap. These muscles were exposed only when gentle probing to identify the muscle origin of the afferents was performed. For EMG recording, pairs of enamel-coated copper wires Ž100 mm diameter, 10 mm spacing, 1.5 mm tip bared. were inserted into the anterior part of the left deep masseter muscle. Three screws were inserted into the dorsal aspect of the skull to mount a stereotaxic frame. For recording muscle spindle units from MesV, the dura was exposed at the occipital area centered at 14 mm posterior to bregma and 2 mm left of the midline. In 18 rabbits, conduction velocity was measured. A branch of masseter nerve was stimulated by the bipolar electrode set at the position where the masseter nerve fibers run under the zygomatic arch, a bone which was removed as carefully as possible. The electrode was covered with Vaseline and a 0.05 ms 70–120 mA square pulse was delivered. 2.2. Unit recording A glass-coated elsiloy microelectrode with impedance 1–3 MV at 1 kHz was inserted vertically through the left cerebral cortex towards the MesV. During the recording, anesthesia Žhalothane, 1.5%, 2 lrmin. was maintained at
such a level that neither a corneal reflex nor a withdrawal reflex from pinching occurred. In addition, neither spontaneous muscle activity nor a jaw-closing reflex resulting from jaw opening was observed. This was done to ensure that the masticatory muscles would be relaxed avoiding loading muscle spindle afferents by concomitant extrafusal muscle contraction and thereby disturbing jaw movement. 2.3. Response to muscle stretch Mechanical stimulation was applied by means of a dual mode mechanical controller attached to the connector in the mentum. The jaw movement applied in the present study was performed in the ramp-and-hold stretch manner. The initial jaw position, at which upper and lower teeth did not contact, was 2.58 from the fully closed position. The center of rotation of the lower jaw was referred to in the report from Weijs et al. w43x. The holding phase lasted 1.5 s and the cycle was repeated every 6 s. Stretch and release speed were the same within each ramp stretch and were changed as required. For each unit, three types of parameters of the response to muscle stretch were defined as follows: Ž1. initial burst ŽIB. was defined as the high instantaneous frequency of firing at the start of the muscle stretch; Ž2. negative adaptation ŽNA. was defined as the abrupt fall in frequency at the beginning of the hold phase, in case that the instantaneous frequency of the first impulse at the beginning of hold phase was lower than that of following impulses during subsequent 100 ms; and Ž3. the firing during release ŽFDR. phase. The dynamic and static responsiveness to muscle stretches were calculated as follows. Two basic measurements were calculated from five muscle stretches and the averages were used in the analysis ŽFig. 1.. The dynamic index ŽDI. was defined as the difference between the mean frequency obtained from the 0.05 s period preceding the end of the dynamic phase and the frequency obtained from the 0.4 s to 0.6 s period after the end of the dynamic phase. The static difference ŽSD. was defined as the difference between the mean frequency obtained from 0.4 s to 0.6 s period after the end of the dynamic phase and the 0.5 s period preceding the ramp stretch onset. Dynamic responsiveness was evaluated by both the DI and the velocity sensitivity ŽVS.. The VS was determined from the slope of the regression line obtained from the linear relationships between the five stretch rates Ž6, 12, 15, 20 and 308rs, with 38 amplitude. and the DIs at each stretch rate. The static responsiveness was evaluated by the position sensitivity ŽPS.. The SD at five stretch amplitudes Ž1, 1.8, 3, 4.8 and 68, at 128rs. was measured and the slope of the regression line between the SDs and the stretch amplitudes was defined as the PS. Although the head of the animal was fixed as firmly as possible, the unitary recordings sometimes failed when the stretches with the fast rates Ž208rs and more. or the large amplitudes Ž4.88 and more. were applied. That is the reason why
T. Kato et al.r Brain Research 833 (1999) 1–9
3
Fig. 1. Unit response to jaw opening at 128rs with 38 amplitude. Top trace: unit discharges. Middle trace: instantaneous frequency. Bottom trace: vertical jaw movement. Upward arrow indicates the direction of jaw opening. DI: dynamic index. SD: static difference.
the number of units stated in the results varies. Discharge variability was defined as the coefficient of variation ŽŽS.D.rmean interspike intervals.100. during the last 500 ms in the hold phase of the stretch with a 38 amplitude.
discrimination and calculation were performed using Spike2 software ŽCambridge Electronic Design..
2.4. Conduction Õelocity of the spindle afferents
To compare two groups, the null hypothesis that each variable had a normal distribution was initially tested by a chi-square test. Then, the null hypothesis that the variance had homogeneity was tested by an F-test. If the two null hypotheses were not rejected, the difference between two groups was tested by a two sample t-test. When either or none of the above two null hypotheses were not rejected, the difference was tested by a Mann–Whitney test. In a comparison among groups, the difference was tested by a Kruskal–Wallis test. Results are presented as mean " S.D. In some cases, to test the assumption that the frequency distribution of quantitative data consists of two subpopulations, data were analyzed by the statistical method of Hald w12x. Mean values and S.D.s for the two possible subpopulations were defined and presented as mean " S.D. This method was used the previous reports w34,39x.
The conduction velocity was calculated from the latency of the evoked action potential and the conduction distance. The distances between the stimulating site to the MesV were measured after post-fixation. 2.5. Histology At the end of the experiments, the animals were killed by an overdose of pentobarbital sodium administered intravenously. Recording sites were marked by passing a negative current of 30 mA through the recording electrode for 10 s. The animals were perfused with saline followed by Formalin Ž10%.. After post-fixation, the brain was removed from the skull and frozen sections were cut in the frontal plane and stained with Cresyl violet. Recording sites were identified within MesV under light microscope.
2.7. Statistical analysis
3. Results 2.6. Data analysis 3.1. Identification of the spindle afferents The unitary discharge, the jaw movements and EMG activity were simultaneously recorded on an 8-channel DAT data recorder ŽPC-208A, Sony-Magnescale, Tokyo, Japan.. The data was replayed and fed into a personal computer ŽDell Optiplex XMT 5133, Dell Computer, Round Rock, TX, USA. with CED 1401 plus interface ŽCambridge Electronic Design, Cambridge, UK.. Spike
Responses of 168 units were recorded from the MesV. The location of the units was confirmed histologically. These units responded to the jaw opening, but did not respond to pressure applied to the teeth, gingiva or vibrissae. Furthermore, the first-order neurons of golgi tendon organs are located in Gasserian ganglia rather than in the
4
T. Kato et al.r Brain Research 833 (1999) 1–9
MesV w21x. Accordingly, all units recorded here were muscle spindle units. The muscle origins of 156 units were identified by gentle muscle probing. Ninety-seven units innervated in the masseter muscle, 19 in the medial pterygoid muscle and five in the temporal muscle. Another 35 units did not respond to muscle probing but did respond to pushing on the eyeball. Because the posterolateral bone of the orbital fossa does not exist in the rabbit, it is assumed that by pushing on the eye-ball some soft tissue around the eye-ball indirectly pushed the deep parts of the masseter muscles or the pterygoid muscles and the spindles were stretched. Most of the units innervating the masseter muscle responded to medial jaw shift and those in the medial pterygoid muscle responded to lateral jaw shift. There were no significant difference in all measurements calculated in the present study among the muscles of origin Ž p ) 0.05, Kruskal–Wallis test.. 3.2. Firing pattern The response to ramp-and-hold stretches was investigated with respect to three qualitative parameters, i.e., IB, NA, and FDR. These parameters are thought to be useful in discriminating between primary and secondary muscle spindle endings w4,35,36x. Among the three parameters, IB and FDR were more related to the rate of the change in muscle length than NA. Fig. 2 illustrates this relationship.
The unit in Fig. 2A did not show an IB in response to stretch at 128rs while an IB was present at 308rs. On the other hand, the unit in Fig. 2B showed FDR at the velocity of release in 128rs, but not FDR at 308rs. Within the range of the velocities in the present study Ž6, 12, 15, 20 and 308rs., the threshold of the velocity of stretch or release at which an IB appeared or the FDR disappeared varied among the units. However, units which showed NA in their response to stretch at 128rs always showed NA at faster velocities than 128rs, and vice versa ŽFig. 2.. In the case when the much faster velocity of stretch and release Ž608rs. was applied to 62 units, units without NA at 308rs never showed NA at 608rs. As a result, the decision whether each unit has NA or not was made most confidently among three parameters, although it is known that the characteristics of these three parameters are dependent on the rate of change of muscle length w14,19,23,33x. Furthermore, there was evidence that NA could be a reliable criterion w4,14,15x. Therefore, all recorded units were divided into two groups on the basis of the existence of NA at the velocity of stretch of 128rs. Out of 168 units, 61 units showed NA ŽNAŽq. units. and 107 units did not ŽNAŽy. units. at the velocity of 128rs. Table 1 shows the number of units with an IB and FDR at two velocities Ž12 and 308rs. and compares NAŽq. and NAŽy. units. At both velocities, NAŽq. units were likely to show more IB and less FDR than NAŽy. units Žin all cases: p - 0.001, chi-square test for independence..
Fig. 2. Response patterns of two units at two different velocities of stretch Ž128rs and 308rs.. A,B: Response of an NAŽq. unit. Although NA existed at both velocities of stretch, an IB was not present at 128rs ŽA. but was at 308rs ŽB.. C,D: Response of an NAŽy. unit. FDR was observed at 128rs ŽC. and disappeared at 308rs ŽD.. Within the range of the velocities of stretch examined in the present study, NA never appeared in this type of unit. Downward arrow: IB, upward open arrow: NA, asterisk: FDR.
T. Kato et al.r Brain Research 833 (1999) 1–9 Table 1 Contingency table of firing patterns at 128rs and 308rs 128rs
IB FDR
Žq. Žy. Žq. Žy.
308rs
NAŽq. Ž ns61.
NAŽy. Ž ns107.
NAŽq. Ž ns 41.
NAŽy. Ž ns65.
13 48 8 53
3 104 69 38
28 13 0 41
7 58 22 43
IB was more frequent in the response of NAŽq. units than NAŽy. units and a higher proportion of NAŽq. units showed IB at both velocities. In contrast, FDR was more frequent in the response of NAŽq. units than NAŽy. units at both velocities.
3.3. QuantitatiÕe analysis The distributions of the DIs for 168 units at 128rs and for 106 units at 308rs are shown in Fig. 3A and B. The
5
DIs were 49.5 " 24.3 impulsesrs for 128rs and 70.8 " 33.5 impulsesrs for 308rs. The DIs at 128rs for 61 NAŽq. units Ž69.2 " 20.4 impulsesrs. were significantly higher Ž p - 0.001; two sample t-test. than those for 107 NAŽy. units Ž38.3 " 18.6 impulsesrs.. A significant difference in DIs at 308rs between the two groups was also observed ŽNAŽq. units: 101.0 " 27.9 impulsesrs, NAŽy. units: 51.8 " 20.2 impulsesrs, p - 0.001, Mann–Whitney U-test.. The distributions of DIs at the two velocities of stretch were not clearly bimodal and there was a large overlap between the two groups. However, the results from Hald’s method revealed that at each velocity of stretch the distribution of DI consisted of two subpopulations. In Fig. 3C and D, two curves of two subpolulations defined by Hald’s method ŽC: 74.1 " 21.1 impulsesrs and 36.1 " 16.7 impulsesrs, D: 115.4 " 34.1 impulsesrs and 51.6 " 18.8 impulsesrs. are shown. In both C and D, it is apparent that
Fig. 3. Histograms of the DI calculated from the responses of 168 units at 128rs ŽA. and from 106 of those units at 308rs ŽB.. Filled bars indicate NAŽq. units and open bars indicate NAŽy. units. At both velocities of stretch, NAŽq. units had a significantly higher DI than those of NAŽy. units ŽA: p - 0.001, two sample t-test, B: p - 0.001, Mann–Whitney U-test.. In C and D, histograms of total distributions of the DI corresponding to A and B, respectively, are shown. Two curves in each histogram indicate frequency distributions of two subpopulations calculated by the method of Hald.
6
T. Kato et al.r Brain Research 833 (1999) 1–9
The VS was determined to evaluate the sensitivity of 106 units to the change of the five different velocities of stretch. VS was distributed mainly from 0.07 impulsesrdegree to 2.63 impulsesrdegree with a maximum of 4.19 impulsesrdegree ŽFig. 4.. The VS of 41 NAŽq. units was 1.85 " 0.55 impulsesrdegree and that of 65 NAŽy. units was 0.83 " 0.41 impulsesrdegree. The VS of the NAŽq. group was significantly larger than that of the NAŽy. group Ž p - 0.001, Mann–Whitney U-test.. The result obtain from Hald’s method is shown in Fig. 4B. Although the frequency distribution of NAŽq. units Žfilled bars. was not normally distributed because of one extreme data point, a correlation between the two curves Ž2.23 " 0.74 impulsesrdegree and 0.81 " 0.5 impulsesrdegree. and the two actual distributions was observed. The VS in the present study was lower than in the masseter muscle of the cat with the mean 30 impulsesrdegree w16x. PS was determined to evaluate the sensitivity of units to the change of the amplitudes of the jaw opening. The
Fig. 4. A: Histogram of the VSs calculated from 106 units ŽA.. Filled bars indicate NAŽq. units Ž ns 41. and open bars indicate NAŽy. units Ž ns65.. VS of NAŽq. units was significantly higher than NAŽy. units Ž p- 0.001, Mann–Whitney U-test.. In B, histograms of total distributions of the VS are shown. Two curves indicate frequency distributions of two subpopulations calculated by the method of Hald.
there is much correspondence with the actual frequency distributions of two groups defined as NAŽq. units Žfilled bars. and NAŽy. units Žopen bars..
Fig. 5. Histogram of the PSs calculated from 59 units. Filled bars indicate NAŽq. units Ž ns 24. and open bars indicate NAŽy. units Ž ns 35.. There was no significant difference between the two types of unit groups.
Fig. 6. A: Histogram of the discharge variability calculated from 168 units. Filled bars indicate NAŽq. units Ž ns61. and open bars indicate NAŽy. units Ž ns107.. The discharge variability of NAŽq. units was significantly higher than NAŽy. units Ž p- 0.001, Mann–Whitney Utest.. In B, histograms of total distributions of the discharge variability are shown. Two curves indicate frequency distributions of two subpopulations calculated by the method of Hald. Bin width of the histogram in B was defined as 3 to increase the number of data points for this calculation.
T. Kato et al.r Brain Research 833 (1999) 1–9
distribution of PS obtained from 59 units is shown in Fig. 5. The mean PS calculated from these units was 16.2 " 5.6 impulsesrs per degree, 17.2 " 6.3 impulsesrs per degree for 24 NAŽq. units and 15.5 " 5.2 impulsesrs per degree for 35 NAŽy. units. There was no significant difference between the two groups. PS in the present study was much higher than that found in cats where a mean of 7.55 impulsesrs per degree was found w16x. Discharge variability for 168 units was determined to compare the regularity of firing during the hold phase between the two groups. The mean discharge variability obtained from 61 NAŽq. units was 23.0 " 9.9 and that of 107 NAŽy. units was 8.2 " 4.3. NAŽy. units discharged significantly Ž p - 0.001, Mann–Whitney U-test. more regularly than NAŽq. units ŽFig. 6.. Two subpopulations determined by the Hald’s method Ž27.0 " 11.5 and 7.7 " 3.8. had a good correlation with the actual two populations of NAŽq. and NAŽy. units ŽFig. 6B.. Some NAŽq. units fired with a small number of spikes during the last 500 ms of the hold phase, which made their discharge variability large. However, even when this source of variability is taken into account, the discharge variability was much larger in the present study than in results reported elsewhere w7,24,31x. Residual fusimotor activity may also be attributed to relatively large discharge variability and PS. Thirty-seven units that followed a repetitive electrical stimulation to the masseter nerve at 333 Hz or above were identified as masseter spindle units. These units responded to muscle stretch and were located in the MesV. Conduction velocities of 37 units ranged from 17.9 mrs to 54.5 mrs with a mean value of 40.0 " 9.2 mrs. As shown in Fig. 7, conduction velocities of the two unit groups heavily overlapped; however, the conduction velocity of 17 NAŽq. units Ž45.3 " 5.1 mrs. was significantly faster than that of 20 NAŽy. units Ž35.5 " 9.5 mrs. Ž p - 0.001, Mann– Whitney U-test..
Fig. 7. Histogram of the conduction velocities calculated from 37 units innervating the masseter muscle.
7
4. Discussion The results revealed that NAŽq. and NAŽy. units had different response characteristics. NAŽq. units showed an IB at the start of stretch more frequently and discharged during the release phase less frequently than NAŽy. units. Dynamic responsiveness, assessed by measuring DI, VS, and discharge variability were higher in NAŽq. units than in NAŽy. units. In the hind limb muscle of the cat, physiological classification of muscle spindle afferents has been employed on the basis of conduction velocity because the distribution of conduction velocities is clearly bimodal. The primary endings are formed by group Ia afferents with larger axon diameters and higher conduction velocities, and the secondary endings by group II afferents with smaller axon diameters and lower conduction velocities w23x. The distribution of conduction velocities of the muscle spindle afferents in the hind limb of the rabbit was also bimodal w10,28x. On the other hand, in masticatory muscles of the cat, the distributions of both axon diameters and conduction velocities of muscle spindle afferents are not clearly bimodal w2,16,25x. In the present study, the distribution of conduction velocities of jaw-muscle spindle afferents in rabbits was not bimodal, either, although there was a significant difference between NAŽq. and NAŽy. units. Therefore, like in cats, it is impossible to classify the muscle spindle afferents in the masticatory muscle of the rabbit on the basis of the conduction velocity. Previous studies revealed that the response to the rampand-hold stretch was different between primary and secondary endings w4,8,23,34,35x. The primary ending shows a high frequency of discharge at the start of the dynamic phase ŽIB., an abrupt fall of discharge at the start of the hold phase ŽNA., and is likely to fall silent during the release phase. Classification based on these response patterns has been reported in some studies w4,35x, but not in the jaw muscles. Cheney and Preston w4x reported that, in the soleus muscle of the baboon, the classification based on NA was consistent with that of the conduction velocity of the afferent. Schafer ¨ w34x reported that in the cat hind limb the afferents with conduction velocity ranging from 80 to 105 mrs showed NA in response to ramp stretch. Hunt and Ottoson w14x recorded the afferent response to a ramp-and-hold stretch from individual axons with primary and secondary endings in an isolated muscle spindle preparation. They found that the rapid fall of the receptor potential at the end of the dynamic phase was evident at a high velocity of stretch only in the primary endings. The rapid fall of the receptor potential could cause the abrupt fall in discharge, NA w15x. These studies suggest that NA may be a useful criteria to classify muscle spindle endings. Since the appearance or amplitude of the IB is dependent on the stiffness of intrafusal muscle fibers w30x, Scott w35x tested the classification by using the IB or combined criteria with the IB, NA and FDR in the cat. They condi-
8
T. Kato et al.r Brain Research 833 (1999) 1–9
tioned the parent muscle to enhance the amplitude of the IB by electrical stimulation to the efferents. They reported that this conditioning procedure improved the discrimination between the two types of afferents. In the present study, conditioning procedures were not performed because it was difficult to stimulate the efferents innervating all the masticatory muscles. This might have caused fewer units with IB in their responses to stretch than in the report by Scott w35x. The distributions of quantitative measurements were not bimodal and there was a large overlap between the two groups. This result seems to be common in the jaw muscles w4,16,39x, and is attributed to the histological differences between the jaw and limb muscles or to the difference in the experimental condition between the present and previous studies. Compared with the cat’s hind limb muscle spindle, there was a high proportion of intermediate spindles in the jaw-muscle spindle in the cat w16,39x. A high proportion of intermediate muscle spindles is related to the existence of complexes of spindles and tandem spindles w21x, and to the various innervation patterns to those spindles w3,32x. In the jaw muscles of the rabbits, however, spindles exist more simply than in the cat w32x. Spindles usually occurred singly, rarely as partially fused pairs, with one bag1, one bag2, and two to seven chain fibers. This histological study may result in less overlap between measurements of the two groups in the rabbit than in the cat. Other potential factors contributing to the large overlap may be various innervation patterns like branching of jaw-muscle spindle afferents outside the capsule w17x and separate innervation to bag1 and bag2 fibers w3x. In addition, the experimental conditions used in this study vs. previous studies should be considered. In rabbits, unitary recordings often failed when the stretch with large amplitude was applied because fixation of the head was less rigid than in cats. Therefore, in the present study, the muscle was stretched at an intermediate amplitude. It was reported that some of the primary units did not show NA in other muscles, for example, the baboon w19x, human w11x, and cat neck muscles w31x. In the study by Edin and Vallbo w11x, primary afferents lacked NA more frequently in humans than in the baboon. They assumed that their result was owing to the fact that the muscles were at an intermediate length during ramp stretch. To help relate the results from this study to the experiments reported previously w13,27x, we chose halothane anesthesia. Pages and Proske w29x suggested that halothane might have a direct action on the receptor, because the dynamic response of primary endings was enhanced by halothane inhalation under deep anesthesia with de-efferentation. It was shown that the effects of electrical stimulation in the different central structures to the responses of muscle spindle endings were modulated by increasing the depth of halothane anaesthesia w40–42x. They did not discuss specific fusimotor activity caused by
halothane anaesthesia. From their results, it is assumed that halothane anesthesia may have had an effect on fusimotor activity by modulating the central nervous system. Under the de-efferented condition, discharge variability of muscle spindle endings in different muscles was generally very low w7,24,31x. It was increased strongly by static fusimotor activity w5,24x. Since discharge variability in this study was in a higher range than those studies with de-efferented or fusimotor-suppressed conditions w7,24,31x, residual static fusimotor activity was likely present in our preparation. It remains unclear how the interaction of central static effects with peripheral dynamic effects produce changes in fusimotor activity and thereby affect the sensitivity of muscle spindle endings. However, since these fusimotor activities modulate the response characteristics of muscle spindle afferents as described previously w5,23x, they could affect the value of quantitative data and then contribute to the large overlap between the two groups. Despite the presence of some confounding factors for classification as discussed above, the present results clearly revealed that the two types of units, divided with respect to the existence of NA, had different response characteristics to muscle stretch. It could be inferred that the units with NA may be regarded as primary endings and units without NA as secondary endings. In our previous report, it was observed that there might be a difference in the response of muscle spindle endings to the different hardness of the test materials used during CRJMs under halothane anesthesia w27x. During mastication in awake animals or during reflexively-evoked jaw movements, different behaviors between dynamic and non-dynamic types of units were observed w6,22,37,38x. Morphological studies revealed that the dynamic–sensitive muscle spindle afferents in the cat and rat jaw muscles project differently to the trigeminal motor nucleus and surrounding areas from the non-dynamic muscle spindle afferents Žrat w9x; cat w18x.. Sensory inputs from jaw-muscle spindles are thought to be importantly involved in regulating masticatory force w13x. However, behavior of the jaw-muscle spindle endings during CRJMs has not been clearly demonstrated and the role of proprioceptive inputs to control masticatory force during CRJMs remains uncertain. In consideration of the studies mentioned above, it is assumed that the two types of unit groups observed in the present study play different roles in the regulation of masticatory force. References w1x K. Appenteng, Jaw muscle spindle and their central connections, in: A. Taylor ŽEd.., Neurophysiology of the Jaws and Teeth, Macmillan, London, 1990, pp. 247–255. w2x K. Appenteng, T. Morimoto, A. Taylor, Fusimotor activity in masseter nerve of the cat during reflex movements, J. Physiol. 305 Ž1980. 415–431. w3x R.W. Banks, D. Barker, H.H. Saed, M.J. Stacey, Innervation of muscle spindles in rat deep masseter, J. Physiol. 406 Ž1988. 161P. w4x P.D. Cheney, J.B. Preston, Classification and response character-
T. Kato et al.r Brain Research 833 (1999) 1–9
w5x w6x
w7x
w8x
w9x
w10x
w11x w12x w13x
w14x
w15x
w16x
w17x
w18x
w19x w20x
w21x
w22x
w23x
istics of muscle spindle afferents in the primate, J. Neurophysiol. 39 Ž1976. 1–8. P.D. Cheney, J.B. Preston, Classification of fusimotor fibers in the primate, J. Neurophysiol. 39 Ž1976. 9–19. F.W.J. Cody, L.M. Harrison, A. Taylor, Analysis of the activity of jaw-closing muscles during normal movements in the cat, J. Physiol. 252 Ž1975. 565–592. F.W.J. Cody, R.W.H. Lee, A.A. Taylor, Functional analysis of the components of the mesencephalic nucleus of the fifth nerve in the cat, J. Physiol. 226 Ž1972. 249–261. S. Cooper, The responses of the primary and secondary endings of muscle spindles with intact motor innervation during applied stretch, J. Exp. Physiol. 46 Ž1961. 383–398. D. Dessem, R. Donga, P. Luo, Primary- and secondary-like jawmuscle spindle afferents have characteristic topographic distribution, J. Neurophysiol. 77 Ž1997. 2925–2944. K. Diete-Spiff, The discharge pattern of muscle spindles of the rabbit on activation of intrafusal muscle fibres, J. Physiol. 159 Ž1961. 282–296. ˚ Vallbo, Dynamic response of human muscle spindle B.B. Edin, A.B. afferents to stretch, J. Neurophysiol. 63 Ž1990. 1297–1306. A. Hald, Statistical Theory with Engineering Applications, Wiley, New York, 1952. O. Hidaka, T. Morimoto, Y. Masuda, T. Kato, R. Matsuo, T. Inoue, M. Kobayashi, K. Takada, Regulation of masticatory force during cortically induced rhythmic jaw movements in the anesthetized rabbit, J. Neurophysiol. 77 Ž1997. 3168–3179. C.C. Hunt, D. Ottoson, Impulse acitivity and receptor potential of primary and secondary endings of isolated mammalian muscle spindles, J. Physiol. 252 Ž1975. 259–281. C.C. Hunt, R.S. Wilkinson, Y. Fukami, Ionic basis of the receptor potential in primary endings of mammalian muscle spindles, J. Gen. Physiol. 71 Ž1978. 683–698. H. Inoue, T. Morimoto, Y. Kawamura, Response characteristics and classification of muscle spindles of the masseter muscle in the cat, Exp. Neurol. 74 Ž1981. 548–560. T. Kato, Y. Kawamura, T. Morimoto, Branching of muscle spindle afferents of jaw closing muscles in the cat, J. Physiol. 323 Ž1982. 483–495. H. Kishimoto, Y.C. Bae, A. Yoshida, M. Moritani, M. Takemura, S. Nakagawa, Y. Nagase, T. Wada, B.J. Sessle, Y. Shigenaga, Central distribution of synaptic contacts of primary and secondary jaw muscle spindle afferents in the trigeminal motor nucleus of the cat, J. Comp. Neurol. 391 Ž1998. 50–63. T.H. Koeze, Muscle spindle afferent studies in the baboon, J. Physiol. 229 Ž1973. 297–317. G. Lavigne, J.S. Kim, C. Valiquette, J.P. Lund, Evidence that periodontal pressoreceptors provide positive feedback to jaw closing muscles during mastication, J. Neurophysiol. 58 Ž1987. 342–358. J.P. Lund, F.J.R. Richmond, C. Touloumis, Y. Patry, Y. Lamarre, The distribution of Golgi tendon organs and muscle spindles in masseter and temporalis muscles of the cat, Neuroscience 3 Ž1978. 259–270. Y. Masuda, T. Morimoto, O. Hidaka, T. Kato, R. Matsuo, T. Inoue, M. Kobayashi, A. Taylor, Modulation of jaw muscle spindle discharge during mastication in the rabbit, J. Neurophysiol. 77 Ž1997. 2227–2231. P.B.C. Matthews, Mammalian Muscle Receptors and Their Central Actions, Edward Arnold, London, 1972.
9
w24x P.B.C. Matthews, R.B. Stein, The regularity of primary and secondary muscle spindle afferent discharges, J. Physiol. 202 Ž1969. 59–82. w25x T. Morimoto, H. Inoue, Y. Kawamura, Diameter spectra of sensory and motor fibers in nerves to jaw-closing and jaw-opening muscles in the cat, Jpn. J. Physiol. 33 Ž1982. 171–182. w26x T. Morimoto, T. Inoue, Y. Masuda, T. Nagashima, Sensory components facilitating jaw-closing muscle activities in the rabbit, Exp. Brain Res. 76 Ž1989. 424–440. w27x T. Morimoto, O. Nakamura, K. Ogata, Z.J. Liu, R. Matsuo, T. Inoue, Y. Masuda, O. Saito, J. Mizuno, T. Kato, Autoregulation of masticatory force in the anesthetized rabbit, in: T. Morimoto, T. Matsuya, K. Takada ŽEds.., Brain and Oral Functions. Elsevier, Amsterdam, 1995, pp. 115–124. w28x P.R. Murphy, The patterns of discharge of static and dynamic g motoneurons in the decerebrate rabbit, J. Physiol. 333 Ž1982. 29–37. w29x B. Pages, U. Proske, Effect of halothane anesthesia on responses from primary endings of muscle spindles in the cat, Exp. Neurol. 28 Ž1970. 393–402. w30x U. Proske, G.J. Stuart, The initial burst of impulses in responses of toad muscle spindles during stretch, J. Physiol. 368 Ž1985. 1–17. w31x F.J.R. Richmond, V.C. Abrahams, Physiological properties of muscle spindles in dorsal neck muscles of the cat, J. Neurophysiol. 42 Ž1979. 604–617. w32x A. Rowlerson, F. Mascarello, D. Barker, H. Saed, Muscle-spindle distribution in relation to the fibre-type composition of masseter in mammals, J. Anat. 161 Ž1988. 37–60. w33x S.S. Schafer, ¨ The acceleration response of a primary muscle-spindle ending to ramp stretch of the extrafusal muscle, Experientia 23 Ž1967. 1027. w34x S.S. Schafer, Regularity in the generation of discharge patterns by ¨ primary and secondary muscle spindle afferents, as recorded under a ramp-and-hold stretch, Exp. Brain Res. 102 Ž1994. 198–209. w35x J.J.A. Scott, Classification of muscle spindle afferents in the peroneus brevis muscle of the cat, Brain Res. 509 Ž1990. 62–70. w36x J.J.A. Scott, Response of Ia afferent axons from muscle spindles lacking a bag1 intrafusal muscle fibre, Brain Res. 543 Ž1991. 97–101. w37x A. Taylor, F.W.J. Cody, Jaw muscle spindle activity in the cat during normal movements of eating and drinking, Brain Res. 71 Ž1974. 523–530. w38x A. Taylor, O. Hidaka, R. Durbaba, P.H. Ellaway, Fusimotor influence on jaw muscle spindle activity during swallowing-related movements in the cat, J. Physiol. 503 Ž1997. 157–167. w39x A. Taylor, R. Durbaba, J.F. Rodgers, The classification of afferents from muscle spindles of the jaw-closing muscles of the cat, J. Physiol. 456 Ž1992. 609–628. w40x J.P. Vedel, Effets des stimulations du noyau caude´ sur l’activite´ des terminaisons fusoriales primaires et secondaires du muscle soleaire, ´ Exp. Brain Res. 22 Ž1975. 113–128. w41x J.P. Vedel, J. Mouillac-Baudevin, Controle ˆ de l’activite´ des fibres fusimotrices dynamiques et statiques par la formation reticulee ´ ´ mesencephalique chez le chat, Exp. Brain Res. 9 Ž1969. 307–324. ´ ´ w42x J.P. Vedel, J. Mouillac-Baudevin, Controle ˆ pyramidal de l’activite´ des fibres fusimotrices dynamiques et statiques chez la chat, Exp. Brain Res. 10 Ž1970. 39–63. w43x W.A. Weijs, J.A.M. Korfage, G.J. Langenbach, The functional significance of the position of the centre of rotation for jaw opening and closing in the rabbit, J. Anat. 162 Ž1989. 133–148.