Properties of rhythmically active reticular neurons around the trigeminal motor nucleus during fictive masticaton in the rat

Neuroscience Research, 14 (1992) 275-294

275

© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0168-0102/92/$05.00 NEURES 00560

Properties of rhythmically active reticular neurons around the trigeminal motor nucleus during fictive masticaton in the rat Tomio Inoue, Yuji Masuda, Tadashi Nagashima, Kenji Yoshikawa and Toshifumi Morimoto Department of Oral Physiology, Osaka University, Faculty of Dentistry, 1-8, Yamadaoka, Suita, Osaka 565, Japan (Received 9 January 1992; Accepted 8 June 1992)

Key words: Sensory; Mastication; Motoneuron; Premotor neuron; Reticular formation; Rhythm; Supratrigeminal nucleus

SUMMARY Response properties of the neurons in the reticular formation around the trigeminal motor nucleus (MoV) were examined during cortically-induced fictive mastication (CIFM) in anesthetized and immobilized rats. Forty-three neurons were rhythmically active (RA neurons) during CIFM, most of which were located in the supratrigeminal nucleus and the reticular formation medial to the oral spinal trigeminal nucleus. The firing frequency of 36 of the RA neurons was modulated in the same rhythm as that of masseteric or digastric nerve activities during CIFM. We divided these neurons into four groups according to the phase of activation: sixteen neurons fired mainly in the phase of masseteric activity (type 1), 11 fired in the transition phase from masseteric activity to digastric activity (type 2), 5 fired in the phase of digastric activity (type 3) and 4 fired in the transition phase from digastric activity to masseteric activity (type 4). Thirty-nine (91%) of the 43 RA neurons responded to at least one of the tested peripheral stimuli. The responses were mostly excitatory but inhibitory responses were sometimes obtained, especially for types-1 and 2 neurons. RA neurons in the reticular formation medial to the oral spinal trigeminal nucleus responded to stimulation of inferior alveolar nerve at a shorter latency than RA neurons in the supratrigeminal nucleus. Fifteen (48%) of 31 RA neurons responded to triple-pulse stimulation of the contralateral cortex. In contrast, only 5(26%) of the 19 RA neurons responded to the ipsilateral cortical stimulation. Stimulation of the ipsilateral MoV was performed on 24 RA neurons, of which 9 responded antidromically (A-RA neurons) at latencies of 0.4-1.4 ms. Eight (89%) of the 9 A-RA neurons received peripheral inputs. The spike triggered averaging method was applied to 4 of the 9 A-RA neurons, ad in all cases short latency field potentials were recorded in the MoV. We conclude that trigeminal premotor neurons receive convergence from central and peripheral inputs. This integration can adjust the appropriate level of motoneuronal excitability during mastication.

Correspondence to: Prof. T. Morimoto, Department of Oral Physiology, Osaka University, Faculty of Dentistry, 1-8, Yamadaoka, Suita, Osaka 565, Japan.

Abbreviations: 3Cx, triple-pulse stimulation of cortical masticatory area; A-RA, antidromically-activated and rhythmically active; AT, antidromically-activated; CIFM, cortically-induced fictive mastication; CPG, central pattern generator; DigN, digastric nerve activity; Freq, instantaneous firing frequency; IA, inferior alveolar nerve; IO, infraorbital nerve; LCx, left masticatory area of the cerebral cortex; MaMP, masseteric motoneuron pool; MassN, masseteric nerve activity; MoV, trigeminal motor nucleus; RA, rhythmically active; RCx, right masticatory area of the cerebral cortex; RfmVo, reticular formation medial to the oral spinal trigeminal nucleus; SpVo, oral spinal trigeminal nucleus; STA, spike triggered averaging; SupV, supratrigeminal nucleus.

276 INTRODUCTION

Integration of peripheral sensory inputs with central nervous system commands arc necessary for the control of masticatory force. The loss of sensory inputs from the periodontal receptors ad muscle spindles greatly reduced jaw-closing muscle activities during chewing 14,38. On the other hand, when these receptors were stimulated by applying the test material between the upper and lower molars during the cortically-induced rhythmic jaw movements, the jaw-closing muscle activities were greatly enhanced 28,38. These findings suggest that sensations from the above two kinds of receptors are related to the control of jaw-closing muscle activities and thus the masticatory force through a positive feedback loop. However, the jaw-closing muscles were unable to be activated merely by pressing the upper or lower molars in the anesthetized rabbit, while these muscles were easily activated by application of the test material during rhythmic jaw movements 38. Thus, it is possible that activity in periodontal sensory afferents to the trigeminal motoneurons may be regulated by a central pattern generator (CPG) circuit for mastication such that they are effective only during chewing the material. Under these condition the discharge of these interneurons occurs by activity in the CPG. However, little is known about the central neuronal networks of this feedback loop. The reticular regions around the trigeminal motor nucleus (MoV), including the supratrigeminal nucleus 29,30 (SupV), the intertrigeminal nucleus and the bulbar parvocellular reticular formation are potential areas for interneurons involved in the above neuronal circuit for the following reasons: (1) those regions contain the premotor neurons for the MoV 3,4,6,10,19,20,23,27,32,33,36,41,43,50,57,64,(2) some of the afferent fibers from the jaw-closing muscles terminate in the SupV, the intertrigeminal regions and the bulbar parvocellular reticular formation 9,34,44,45,54,55,while some afferent fibers from the periodontal tissues terminate in the SupV 45,53,56, the intertrigeminal regions, the juxtatrigeminal region and the oral spinal trigeminal nucleus (SpVo) 53,5~,, (3) some of these regions in the pontine and bulbar reticular formation receive descending fibers from the cerebral cortex 26,35,58,67and from the amygdaloid complex 13,24,48.Therefore, trigeminal premotor neurons may integrate activity of peripheral inputs with activity in a central pattern generator to produce the appropriate discharge pattern in trigeminal motoneurons. The aim of the present study was to investigate whether trigeminal premotor neurons are involved in the regulation of masticatory activity. For this purpose, we recorded the activities of these reticular neurons around the MoV, which were rhythmically active during fictive mastication induced by repetitive cortical stimulation (RA neurons), and examined their responses to stimulation of the sensory branches of the trigeminal nerve. We also investigated whether some of these RA neurons could be antidromically activated from the ipsilateral MoV stimulation and thus classified as trigeminal premotor neurons. Preliminary results have been published previously 15,1~,, in abstract form. MATERIALS A N D METHODS

Thirty male Sprague-Dawley rats, weighing 300-400 g, were used. The animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (60 mg/kg) and ketamine hydrochloride (30 mg/kg). Supplemental doses of ketamine hydrochloride (20-30 m g / k g / h ) were given via a cannula placed into the femoral vein to maintain anesthesia at such a level that the animals showed a constant heart rate, no apparent

277 corneal reflex and no spontaneous eye movements during surgery and the experiments. After paralysis, the anesthesia was judged to be effective because a noxious paw-pinch did not elicit any change of heart rate. The trachea was cannulated. The rectal temperature was maintained at 36-38°C with a heating pad and the electrocardiogram was continuously monitored. The animal's head was fixed in the frame of a stereotaxic apparatus.

Stimulation Electrical stimulation was applied to the left infraorbital nerve (IO) and inferior alveolar nerve (IA). For this purpose, IO was exposed in front of the infraorbital foramen. A pair of enamel coated stainless wires (diameter: 50 /xm, tip separation: about 1 mm) were inserted into the nerve by 0.5 cm. To stimulate the inferior alveolar nerve (IA), the mental nerve was first exposed in front of the mental foramen, and then the electrodes similar to those used for IO stimulation were inserted centrally along the nerve trunk by about 1.5 cm. Post mortem observation was made to ensure that the cut ends of the electrodes were located in the nerve proximally to the last lower molar. The nerves were stimulated with single or triple pulses (0.2 ms, 500 Hz) at 1.5 times the threshold of evoking the digastric muscle activity. Kidokoro et al. zz reported that the jaw-opening reflex recorded in the digastric nerve can be evoked by inferior dental nerve stimulation at only slightly suprathreshold of the nerve. Thus, the stimulus strength used in the present study is considered relatively low. The left masseteric nerve was exposed under the zygomatic arch and a Teflon coated platinum iridium fine wire (diameter: 25 /xm) was inserted centrally into the nerve for stimulation. This electrode was also used to record efferent activities. The reference electrode of an enamel coated stainless steel wire (diameter: 200 txm) was implanted into the left masseteric muscle near the masseteric nerve. For stimulation of the cerebral cortex, the right cortical surface was exposed between 2 and 3 mm anterior to bregma and between 5 and 5.5 mm lateral to the midline. A glass-insulated elgiloy electrode ( 3 0 - 4 0 / x m exposed tip, impedance 3 MY2 at 1 kHz) was vertically inserted into the cerebral cortex and fixed at the point where the rhythmic jaw movements could be elicited by continuous monopolar stimulation (0.2 ms, 25 Hz) at an intensity less than 60/xA. The stimulating sites of the present experiment corresponded approximately to the posterior area of Sasamoto et al. 52 The occipital part of the cerebral cortex and the dorsal part of the cerebellum on the left side were exposed for stimulation of the left MoV. A glass-insulated elgiloy electrode ( 5 0 / x m exposed tip, impedance 2 M/2 at 1 kHz) was inserted stereotaxically with a 25 ° anterior-to-posterior inclination into the left masseteric motoneuron pool (MaMP) and fixed at the point where the maximal antidromic field potential was evoked by stimulation of the masseteric nerve. A negative stimulating pulse of 0.1 ms duration which was followed by a positive pulse of 0.1 ms duration, was used for stimulation of the MoV in order to minimize the stimulus artifacts in the records of single unit activities (cf. ref. 5). The exposed brain was covered with liquid paraffin. To reduce brain pulsations and to prevent swelling of the brain, drainage of cerebrospinal fluid was performed through an opening in the dura over the foramen magnum.

Recording In order to exclude the effect of rhythmic sensory inputs during the cortically-induced rhythmic jaw movements, animals were immobilized with pancronium bromide (30-50 m g / k g i.v.) and artificially ventilated. The left digastric nerve was exposed and efferent activities were recorded by use of electrode similar to that inserted into the masseteric

278 nerve. Fictive mastication was monitored with rhythmic activities of the digastric a n d / o r masseteric nerves. Supplementary doses of pancronium bromide (5 rag) were given when E M G activities of masticatory muscles were observed. Glass microelectrodes filled with 2 M NaCI solution saturated with fast green FC} ~ were used for recording neuronal activities in the reticular formation around the MoV. The premotor neurons were identified by either of the following two criteria: one is that the recorded neurons respond antidromically to stimulation of the MoV at an intensity less than 20 /zA and the other is that the positive or negative field potential could bc induced in the MaMP when the spike triggered averaging method (STA)40.(~3 was applied to the same neurons. The responses of the reticular neurons to passive jaw movements were examined by manual depression of the mandible. Jaw movements were monitored with an isotonic transducer connected to the mandible. The single cell activities, field potential activities in the MaMP, the masseteric and digastric nerve activities and the passive jaw movement data were amplified, displayed on an oscilloscope and recorded on magnetic tape. The tape was later replayed and filmed. In addition, data during stable CIFM or at least 5 min after the last stimulation were selected and analyzed with the aid of a signal processor (7T-18, Nihondenki Sanei). Data were digitized at 10 kHz or 100 kHz, and calculation of the instantaneous firing frequency, the pre-trigger analysis, averaging of the nerve and field potential activities were performed. Cortically-related stimulus artifacts were suppressed with the use of the signal processor before performing the averaging of nerve activities. Peristimulus time histograms (PSTHs) were made to 32 consecutive stimuli of IA, IO, RCx and LCx. Inhibitory response was considered to be evoked if the number of spikes of a unit during a certain period after the onset of stimuli had been less than 70% of the number of spikes during a period of the same length before the onset of stimuli in PSTH.

Histology At the end of each experiment, the animals were deeply anesthetized by pentobarbital. The recording sites around the MoV were marked by iontophoretic deposition of dyes from the recording electrodes. The recording and stimulating sites in the MaMP and the stimulating site in the cortex were marked by passing negative current of 3 0 / z A for 10 s through the electrodes. The animals were perfused with saline, followed by 10% formaldehyde. After post fixation, the brain was removed from the skull and sectioned serially at 50 tzm in the frontal plane and stained with cresyl violet. Then the recording and stimulating sites were identified. RESULTS A total of 221 neurons were recorded extracellularly from the reticular formation around the MoV. O f these neurons, 43 showed a rhythmic alteration in firing frequency during CIFM ( R A neurons). The rest were either not rhythmically active or the frequency of their rhythmic alteration was different from the rhythm of the CIFM (non-RA neurons). Twenty of the 221 neurons responded antidromically to stimulation of the MoV (AT neurons). The firing frequency of 36 of the above 43 R A neurons was modulated at the same rhythm as that of masseteric or digastric nerve activities during CIFM. Some neurons were activated almost in accord with the phase of the masseteric or digastric nerve activity. The others were activated in the transition phase of the masseteric and digastric nerve activity. We divided these 36 R A neurons into four

279 TABLE I CLASSIFICATION OF THE RA NEURONS Not all cells were tested in every category of stimulation. Type

n

S p o n t . RCx act.

Type 1 Type 2 Type 3 Type 4 Unclassified Total

E

LCx

Stretch

NR

E

NR

E

IA

I

NR

E

IO I

NR

E

I

NR

16 11 5 4

14 9 2 4

6 1 2 3

4 6 3 0

2 0 2 1

5 5 3 1

6 2 0 0

1 1 0 0

6 8 4 3

7 8 3 3

5 3 0 0

3 0 2 0

7 9 4 3

4 2 0 1

2 0 1 0

7 43

6 35

3 15

3 16

0 5

0 14

3 11

1 3

2 23

3 24

1 9

2 7

2 25

0 7

0 3

groups according to the phase of activation: those firing mainly in the phase of masseteric activity (type 1), those firing in the transition phase from masseteric to digastric activity (type 2), those firing in the phase of digastric activity (type 3) and those firing in the transition phase from digastric to masseteric activity (type 4). The numbers of neurons contained in groups 1-4 were 16, 11, 5 and 4, respectively. The remaining 7 R A neurons could not be classified into any group since recording from neither the masseteric nor digastric nerve was successful. We examined responses to stimulation of the contra- and ipsilateral cerebral cortex, to passive jaw opening and to stimulation of I A and I O with 43 R A neurons, though all neurons were not tested in every category. Thirty-five (81%) of the 43 R A neurons were spontaneously active. Regular and irregular spontaneous activity were observed in 17 and 15 neurons, respectively. Three neurons showed burst type spontaneous activity. The results were summarized in Table I. Type 1 neurons Type 1 neurons were the most numerous of all types (16/36). The response to I A stimulation was studied on 15 neurons and 12 (80%) of them responded to stimulation of I A (Table I). A m o n g these 12 neurons, 7 (59%) neurons were activated at a latency of 2.9-7.6 ms (4.8 _+ 1.9 ms) and the remaining 5 (41%) were depressed from IA stimulation (Table I). The response to IO stimulation was also examined in 13 neurons and 11 (85%) of them responded to I O stimulation (Table I). Seven (64%) of these 11 neurons were activated at a latency of 2.9-24.2 ms (8.6 _+ 7.9 ms) and the remaining 4 (36%) were inhibited by I O stimulation (Table I). Seven (54%) of the 13 type 1 neurons also responded to passive jaw opening (Table I). Six (86%) of them excited, while the remaining one was inhibited (Table I). The above 3 kinds of peripheral stimuli (passive jaw opening, I A and I O stimulation) were applied to 12 of the type 1 neurons; 6 (50%) were activated or inhibited by 3 kinds of stimuli, and 4 (33%) were activated or inhibited by 2 kinds of stimuli. We examined inputs from the cerebral cortex in 10 R A neurons, and six were found to receive excitatory inputs either from the contra- or ipsilaterat cortex. Representative responses of a type 1 neuron are shown in Fig. 1. This unit, recorded in the reticular formation medial to the SpVo (RfmVo), showed rhythmical burst activities during CIFM. The instantaneous firing frequency reached about 430 impulses per s, though it lacked spontaneous discharge (Fig. 1A). When each record of the unit and nerves was s u m m a t e d 10 times by triggering with the rising phase of the masseteric

280

A

46-2 T 1 URit :~':..¢-.~L%~::.:~'T%~4:~:,.%.4.- -_:_~~.:-..J_~ p~

.J~L.

JiLl.

,,Jli...

it~

.......ill.

,. l~ J~l...i I 1mY

1r!'~] i] !T'~,! TI~ I' !",W,~7~I,T'~,,~'~,''~

Freq J'"~ ...... ~...............................

Mass N

ll.k4i ll~J] J ;~il;il~i[tli~lLht,.,,L ii :li"~r!Ir~,l'!rrr'r;!rrr~rr,!~(wrrrr~'!

Dig N

B

:-~-.: - ,:,

4

4:-t!,,,,,'!,!!,

Rex ~----.~k-,--~,--

~¢@lqr'~ • ~':~!1 ~

MQSS N ~ Dig N

~[i Lira.; 1ii,,L~L Ll, i i ~,,, ~_;J~l ,,~ ii,,.i • ~;, iu, I'1= ]"r' l'rlq"qrr Tl'~' "- [1~ rT'r'r,[~-,,":l 2 ° n:v , 200ms

C

J Unit

.,;

~

~o,,,,

~'

~

'

,o ~ ~ . . .

,cx ~ l , o v

~

lore 5oms

Fig. 1. Responses of a type 1 neuron in the RfmVo. A: unit activities at rest (left traces) and during repetitive stimulation of the CMA (right traces). B: cumulative histogram of the unit firing during CIFM with the averaged nerve activities. The records were summated 10 times by triggering with the rising phase of the masseteric nerve activities. Nerve activities were rectified before performing the averaging. The triggering point is indicated by the arrow head. C: responses to trigeminal (IA, IO) and cortical (RCx, LCx) stimulation. See text for abbreviations.

nerve activity, it is clear that the unit started and ceased its firing simultaneously with masseteric activity (Fig. 1B). This unit responded with a burst discharge to IA and IO stimulation at the mean latencies of 3.1 ms (_+0.1 ms, n = 11) and 5.6 ms ( + 0 . 7 ms, n = 7), respectively. It also responded with one or two spikes to triple-pulse stimulation (interpulse interval 2 ms) of 1 Hz applied to the contralateral cortex. The latency of the initial spike was 48.0 ms ( -+ 14.1 ms, n = 10). This unit did not respond to stimulation of the ipsilateral cortex. Fig. 2 shows another example of the responses of type-1 neurons which showed inhibitory responses to peripheral stimulation. This neuron, found in the SupV, dis-

A

B

C

33-2 T 1

T

Unit

~-~ ......

~1

Imv

Unit

s~ll~n, Mass

,~1'

N ~,~¢?m~.~~

5ms

"~~ ~ ~ _ 50ms

Jaw

--\~_~--~,~

~---I

4ram

I'"v ~,ss N ~ l l l l l ~ , o . , 20ms

1s

Fig. 2. Responses of a type ! neuron in the SupV. This neuron was inhibited by stimulation of IA and IO (B) and to passive jaw opening (C). A: cumulative histogram of the unit firing during CIFM with the averaged nerve activities. The records were summated t0 times by triggering with the rising phase of the masseteric nerve activities as indicated by an arrow head. See text for abbreviations.

281 58-2 T2

A Unit

' J lmV

Freq Mass N

Dig N

IIII III l Ill

......

....

~

.........

ihLJJ~"I'hLI~]L'~III~'IL]'J'LWL"I'Id'JI'J~IILI'U'I'L~J]~'I

LlJ'l~"~lJikL

,,.,.,L . . . . . . . . .

].~ ............. ]~i~i~d~i~'~i~d~di~i'~Wj~i~j~i~hlL~,

~_, ......

l~

lIIIIill~pfl~lllll~ll~llllil~llp~irlr~ll~t~l~l~irll~rll~II~T~',r ' ~,v :,-:. : ~ :-:. : :: ~: i ,-J ,o~v

~' "

0.5s

,

B

C

D IA ~

Unit ~

l

1my

Jaw

Unit

__, 0.5s

Mass N

2ms

U .. ~kltlllllBI l l J l i l l / h , nit ~,,,,,,,,~,p,,~,,,,,,..= ~q,~qlll lmV

Jaw

D~gN 50ms

~

1

smm 0.Ss

Fig. 3. Responses of a type 2 neuron in the RfmVo during CIFM (A, B), to IA and IO stimulation (C), to light pressure on the left upper incisor (D), and to passive jaw opening (E). Heavy lines in A, D indicate the period of repetitive simulation of the CMA and the period of application of the pressure on the tooth, respectively. The records in B were obtained after summating 10 times by triggering with the rising phase of the masseteric nerve activities as indicated by an arrow head. Upward deflection of the trace of Jaw in E indicates upward jaw movement. See text for abbreviations.

charged spontaneously at the rate of about 20 impulses per s. During CIFM, the neuron discharged coincident with masseteric activity (Fig. 2A). Firing of this unit was depressed for about 25 ms after IA stimulation and 160 ms after IO stimulation, respectively (Fig. 2B). Furthermore, the unit became silent during passive jaw opening (Fig. 2C). Type 2 neurons

Type 2 neurons were almost as numerous as type 1 neurons (11/36). They received convergence from various peripheral inputs. All type 2 neurons responded to stimulation of both IA and IO (Table I). Eight (73%) of them were activated from IA stimulation at a latency of 1.8-23 ms (8.6 _+ 8.2 ms), while the remaining 3 neurons (27%) were depressed by IA stimulation (Table I). When IO was stimulated, nine (82%) of the 11 neurons were activated at a latency of 1.7-44 ms (9.9 _+ 13.9 ms), while the remaining 2 (18%) were inhibited (Table I). The response to passive jaw movement was tested in 11 neurons and 2 (18%) were found to be excited and one (9%) was inhibited (Table I). Seven neurons were tested with single cortical stimulation but 6 of them did not respond and only one was excited (Table I). Fig. 3 shows an example of the responses of type 2 neurons. This unit, recorded in the RfmVo, fired constantly at about 25 impulses per s at rest (Fig. 3A). During CIFM, the neuron showed rhythmical activities corresponding to the transition phase from masseteric activity to digastric activity (Fig. 3B). The maximum firing rate reached about 380 impulses per s (Fig. 3A the record under the thick bar). This neuron received peripheral

282

47-3T3

A Unit . . . . . . . .

IF . . . .

~ ..... ~

I

!p r ! l / ~ W ~ F ~ ! ~ l ~ r r l n l ~

iTr!~li~!~~!i ~mv

i

Freq ...................

~.:...:.;.,..::;~.:.:?:~:...:.,-~;...,:::~;: _.:L.~......... . I tl|IIIttltll ~ l,Jl;'[ttI1~II|,I ' '] I H

. . . . . . . .

r-,

. . . . . . . . . . . .

"

. . . . . . .

.

........

' .

.

.

.

.

.

:

.

. . . . . . .

.

.

. .

I 800Hz

' [ I. . . . '

-'-r !lS!THlTlt !lt!rrll rF !

' - '

.

r ~

.

.

.

.

.

.

.

.

.

.,, ,o,.,,,

.

200ms

a

C

!

Unit

Rcx

"

Mass N

I0 ~ P ~ / " ~ , t " ~ Dig N

LCx " 5ms

~

~

1

1my 5ms

50ms

Fig. 4. Responses of a type 3 neuron in the RfmVo during CIFM (A, B) and to peripheral and cortical stimulation (C). Thick bars in A indicates the period of repetitive simulation of the CMA. The records in B were obtained after summating 10 times by triggering with the rising phase of the masseteric nerve activities as indicated by an arrow head. See text for abbreviations.

convergence. It responded with two spikes to IA and IO stimulation at the mean latencies of 2.0 ms ( + 0 . 3 ms, n = 7) and 1.7 ms (+0.1, n = 8), respectively (Fig. 3C), Furthermore, the neuron responded to passive jaw opening (Fig. 3E) and also to light pressure on the left upper and lower incisors and the left upper molars (Fig. 3D, the record under heavy lines). Type 3 neurons

Five neurons were classified as type 3 neurons according to the phase of activation during CIFM. Most of them responded to stimulation of IA and IO at latencies of 3.0-3.5 and 2.4-6.2 ms, respectively, and no inhibitory response was evoked (Table I). Three of the 5 type 3 units were activated by cortical stimulation at latencies of 15.6-30.2 ms. No type 3 unit responded to passive jaw opening. An example of the responses of the type 3 neuron, found in the RfmVo, is given in Fig. 4. Although the unit showed low and irregular spontaneous discharges, the discharge rate increased to approximately 700 impulses per s during the phase of digastric activity evoked by repetitive cortical stimulation (Fig. 4A,B). This neuron was activated to both peripheral and cortical stimulation. It responded with a burst discharge to IA stimulation at the mean latency of 3.3 ms ( + 0.3 ms, n = 10) and responded with 2 or 3 spikes to IO stimulation at the mean latency of 6.2 ms (_+0.8 ms, n = 5) (Fig. 4C). This neuron showed a rapidly adapting response to pressure on the ipsilateral maxillary molars. It also responded with two or three spikes to stimulation of the contralateral cerebral cortex at the mean latency of 15.6 ms ( + 1.8 ms, n = 16) though it did not respond to stimulation of the ipsilateral cortex (Fig. 4C).

283 4 9 - 1 T4

A Unit

:ii

Freq

..

:(

I

i

......

i

I

'i

i

.LLLLI I L[ I I., LLL] I I.[.,I,~ [ I I = I L~ I I , I I.~ I

I~1 IIFFIIli111fl I1~[I]11 qlllll I1 I1111lrl I I iij II I~mv "'

. . . .

"

1200Hz

lll,J,i~

[W II.I/[JJWI UJJ llIJJ v llYl]rl ffTITIITT] lffl IF]I 11 o~ ~I.LI.WlliLL.tIIII.tLI,/ILW JL 1 ?TEl I'1"1T1T1Ir]"~'l I Fl'ml ITrT 1I';?F l mo.v

Dig N ...............................

200ms

B

C

! Unit

RCx

~,r,h,Afl~ [~

=o ~r~¢-

Mass N

......

LCx ~

IlmV 5ms

Dig N 50me

Fig. 5. Responses of a type 4 neuron in the RfmVo during CIFM (A, B) and to peripheral and cortical stimulation (C). Left and right traces in A indicate the neuronal activities at rest and during repetitive stimulation of the CMA, respectively. The records in B were obtained after summating 10 times by triggering with the rising phase of the masseteric nerve activities as indicated by an arrow head. See text for abbreviations.

Type 4 neurons Four neurons were classified as type 4. Most of type 4 neurons responded with spikes to stimulation of I A and I O at latencies of 2.3-9.6 and 2.7-3.4 ms, respectively. No inhibitory response was evoked from stimulation of either I A or I O with an exception of one neuron (Table I). All type 4 units were activated by stimulation of contra- or ipsilatera[ cortex at latencies of 9.9-19.6 and 16.0 ms, respectively. No type 4 unit responded to passive jaw opening. A representative response of type 4 neuron, found in the RfmVo, is shown in Fig. 5. This neuron discharged spontaneously at a low and irregular firing rate of 3 - 5 Hz (Fig. 5A). When the contralateral cortex was stimulated repetitively, the firing rate increased at the transition phase from digastric activity to masseteric activity (Fig. 5A,B). The unit responded with a burst discharge to I A and I O stimulation at the m e a n latencies of 2.8 ms (__+0.4 ms, n = 16) and 2.8 ms (_+0.5 ms, n = 9), respectively (Fig. 5B,C). The neuron also responded with one or two spikes to single stimulation of the contralateral cerebral cortex at the mean latency of 19.6 ms (_+ 4.7 ms, n = 16) and with two to three spikes to triple shock stimulation of the ipsilateral cortex at the m e a n latency of 16.0 ms (_+ 3.8 ms, n = 17) (Fig. 5C). It also responded to light touch of the dorsal surface of the tongue. This neuron did not respond to passive jaw opening. Unclassified RA neurons Seven R A neurons were not classified corresponding to the phase of masseteric or digastric activity. We could observe rhythmic E M G activities of masticatory muscles before immobilization and after 2 - 3 h of administration of the paralyzing agent. The rhythmic activity of the 7 neurons was similar to that observed in the E M G activity. Thus, it is likely that the activities of the 7 neurons were modulated by inputs from a CPG. Those neurons also responded to passive jaw opening, to single stimulation of

284 T A B L E I1 CLASSIFICATION

OF THE A-RA NEURONS

N o t all cells w e r e t e s t e d in every c a t e g o r y o f s t i m u l a t i o n .

E ! NR

RCx

LCx

Stretch

IA

IO

I 0 4

2 0 5

2 0 4

3 2 2

5 2 1

E = excitatory; I = inhibitory; NR = no response.

contra-, ipsilateral cortex and to I A a n d / o r summarized in Table I.

IO stimulation. Their responses are

Antidromically-activated neurons (A T neurons) To determine if R A neurons project to the MoV, in 24 R A neurons the response to stimulation of M o V was tested. To minimize the possibility of exciting passing fibers near the MaMP, stimuli at intensity less than 20 /xA were delivered. Nine of them responded antidromically at short latencies of 0.4-14 ms (0.72 _+ 0.36 ms) and they followed up to 100 Hz stimulation ( A - R A neurons). These A - R A neurons were found in each of above 4 types. The numbers of A - R A neurons for types 1, 2, 3 and 4 were 3, 1, 3, and 2, respectively. The remaining 15 R A neurons did not respond antidromically. In addition to the above 9 A - R A neurons, 11 neurons of 63 n o n - R A neurons responded to antidromic stimulation. Most A - R A neurons received various peripheral inputs like other R A neurons. The response to I A stimulation was studied on 7 AoRA neurons and 3 (43%) of them were activated at a latency of 2.8-3.9 ms (Table II). Two (29%) were depressed from IA stimulation. Seven (88%) of 8 A - R A neurons responded to I O stimulation (Table II). Five (63%) were activated at a latency of 2.4-2,9 ms (2.7 + 0.19 ms) and 2 (25%) were inhibited by I O stimulation. When the jaw was manually stretched, 2 (33%) of the 6 neurons were excited (Table II). Six neurons were tested with triple-pulse stimulation of cortex and 2 received excitatory inputs either form the contra- or ipsilateral cortex. Fig. 6 shows an example of type 4 A - R A neurons. This is the same example shown in Fig. 5. Antidromic spikes were evoked at a constant latency of 0.6 ms and followed 500 Hz triple shocks to the M o V (Fig. 6A). Furthermore, an antidromic spike was abolished during collision with the preceding spike evoked by I A or I O stimulation (Fig. 6B). These findings suggest that they may be the p r e m o t o r neurons for the trigeminat motoneurons. We further identified this by applying the spike triggered averaging method. For this purpose, the field potential in the M a M P were averaged after spontaneous spike potentials in 4 of the above 9 R A neurons. For all of them, the negative or positive field potentials were induced with latencies of 0.25-1.11 ms. Figure 6C shows such an example. The positive field potential, subsequent to a brief triphasic potential with a latency of 0.43 ms after the onset of spike potentials of the reticular neuron, appeared at a latency of 1.11 ms. The recording and stimulating site i n the M a M P was located 550 txm rostral to caudal pole of the M o V (Fig. 6D), Fig. 7 shows another example of an A - R A neuron. This neuron was classified as type 3 neuron. It responded to I A and I O stimulation at the m e a n latencies of 3.5 ms ( + 0.36 ms, n = 16) and 2.9 ms ( + 0.13 ms, n = 16), respectively, and also to single stimulation of the ipsilateral cortex at the m e a n latency of 30.2 ms ( + 12.7 ms, n = 24). The neuron did not respond to single stimulation of the contralateral cortex (Fig 7A). The antidromic

285 49-1 T4

A MaMP--'~ -~ , , , 1ms MaMP ~ --~,. ~f'~ ~ J ~ ' ~ IlmV 1ms

B IA+ I0+ ~ MaMP ~

"

/'-'~ "----"----~/11mV ;

C

l~,s

MaMP ~

12~v

ilmV

Unit - -

1ms

Fig. 6. Responses of an A-RA neuron. This neuron is the same one shown in Fig. 5. A: antidromic responses to single (upper trace) and 500 Hz triple-pulse (lower trace) stimulation of the left MoV. The filled circles indicate stimulation of the MoV. B: an antidromic spike was abolished in collision with the preceding spikes evoked from IA or IO stimulation. The filled triangles indicate IA or IO stimulation. C: the averaged field potential (5000 sweeps) in the MaMP (upper trace) after the spike potential (lower trace) of the triggering neuron. Positive deflection in field potentials is shown upward in this and following figures. * Recording and stimulation site in the MaMP. See text for abbreviations. spikes w e r e e v o k e d at a c o n s t a n t l a t e n c y o f 0.5 ms a n d f o l l o w e d 500 H z triple shocks to t h e M o V (Fig. 7B). T h e s p i k e was a b o l i s h e d during collision with t h e p r e c e d i n g spike e v o k e d by I A s t i m u l a t i o n (Fig. 7B). W h e n t h e field p o t e n t i a l was s u m m a t e d in the M a M P , a positive field p o t e n t i a l was r e v e a l e d with a l a t e n c y o f 0.25 ms a f t e r the o n s e t o f t h e s p i k e o f t h e r e t i c u l a r n e u r o n (Fig. 7C). T h e r e c o r d i n g a n d s t i m u l a t i n g site in the M a M P was l o c a t e d 1 0 0 0 / , m c a u d a l to r o s t r a l p o l e o f t h e M o V (Fig. 7D). A s d e s c r i b e d above, 15 R A n e u r o n s d i d not r e s p o n d to s t i m u l a t i o n o f t h e M o V a n t i d r o m i c a l l y ( n o n - A T n e u r o n s ) . W e c o m p a r e d r e s p o n s e s o f 10 o f t h e s e n o n - A T but R A n e u r o n s to s t i m u l a t i o n o f the c o n t r a l a t e r a l c e r e b r a l c o r t e x with t h o s e o f 15 o f t h e 20 A T n e u r o n s . O n l y o n e ( 7 % ) A T n e u r o n r e s p o n d e d to s t i m u l a t i o n o f the c o n t r a l a t e r a l cortex, in c o n t r a s t to 6 ( 6 0 % ) n o n - A T n e u r o n s which d i d r e s p o n d . T h e d i f f e r e n c e was r e g a r d e d as significant ( F i s h e r ' s exact p r o b a b i l i t y t e s t , a = 0.014, P < 0.05). O n t h e o t h e r h a n d , only a few o f b o t h A T a n d n o n - A T n e u r o n s r e s p o n d e d to s t i m u l a t i o n o f the i p s i l a t e r a l c o r t e x a n d no significant d i f f e r e n c e was o b s e r v e d .

286 C

A

50-1

T3

I

MaMP

Rcx

......

". . . . . . . . .

,JJl5 ~ , ~ Unit

• cx

--

~ 1 i~

...................

ImV

5ms

B

It lo~v

- -

:~! MaMP~::i~~

1ms

D

MaMP ~i~/~i~i~Dmv Ims

IA+ MaMP

~ ~ I

~-v~j

',~llmV

Fig. 7. Responses of A-RA neuron in the RfmVo. A: responses to stimulation of IA, IO, RCx and LCx. B: antidromic responses to single (upper trace) and 500 Hz triple-pulse (middle trace) stimulation of the left MoV. An antidromic spike was abolished in collision with the preceding spikes evoked from IA stimulation (lower trace). The filled triangles and filled circles indicate stimulation of IA and the MoV, respectively. C: the a v e r a g e d field potential (2000 sweeps) in the MaMP (upper trace) after the spike potential (lower trace) of the neuron. * Recording and stimulation site in the MaMP. See text for abbreviations.

No significant difference was observed between AT and non-AT neurons in response to passive jaw opening and to stimulation of the ipsilateral cortex, IA and IO either. Anatomical localization Fig. 8 shows the recording sites of the RA and AT neurons which were plotted collectively in drawings of the representative histological sections of the brain stem. The locations for representative neurons shown in Figs. 1-7 are all marked in the drawings. Twenty-one RA neurons were located in the SupV, 13 in the RfmVo, 6 in the parabrachial nucleus and the superior cerebellar peduncle and 3 in the paralemniscal

TABLE III LATENCIES OF RESPONSES OF RA NEURONS TO PERIPHERAL AND CORTICAL STIMULATION Values are in ms. Neuron group

SupV

Stimulus

3Cx

IA

IO

3Cx

RfmVo

Minimum Mean Maximum SD n

8.3 17.5 36.8 9.95 6

3.4 5.8 9.6 2.41 8

2.4 6.9 24.2 7.4 8

9.9 22.7 48.0 14.12 6

IA

I0

1.8

1.7

3.1 6. l 1.17 10

3.61 6.2 L59 10

287

Fig. 8. Locations of the R A and A T neurons sampled on the drawings of the representative histological sections separated by 250 ~ m . The neurons presented in the preceding figures are denoted by figure numbers. Filled circles, open circles and small dots indicate A - R A neurons, R A neurons which were not antidromicallyactivated, and A T neurons which were not rhythmically-activated, respectively.

tegmental field. Eleven of the 16 type 1 neurons were located in the SupV, which indicates that type 1 neurons occupied 52% of 21 R A neurons that were located in the SupV. The remaining 3 types of neurons were distributed widely in the SupV and the RfmVo, and no particular topographical arrangement was found for these neurons. The A - R A neurons were scattered in the SupV and RfmVo. Latencies for evoking neuronal discharges were compared between R A neurons in the SupV and the R f m V o (Table III). The mean latency of SupV neurons in response to I A stimulation was 5.8 ms (SD _+ 1.41, n = 8). The mean latency of RfmVo neurons was 3.1 ms (SD _+ 1.17, n = 10). The difference of the latencies was regarded as significant (t-test; t = 3.10, P < 0.01). No significant difference was found between R A neurons in the SupV and the R f m V o in other category of stimulation. DISCUSSION

Identification of premotor neurons for the MoV As described in the Introduction, the trigeminal premotor neurons are located in reticular regions including the SupV, the intertrigerninal region, the parvocellular region and in the SpVo. Previous studies showed that these brainstem regions receive periph-

288 eral sensory inputs from the periodontal mechanoreceptors, muscle spindles and they also receive inputs from higher brain centers including the CMA and the amygdaloid complex. In the present study, 2(} neurons responded antidromically to stimulation ~1 the left MoV in the left SupV and the left reticular formation (RfmVo) medial to the SpVo. These antidromically activated neurons (AT neurons) are probably trigcminal premotor neurons, but fibers passing in and around the MoV could have been activated. Abzug et al. ~ reported that 10/J.m//xA was an estimate of effective spread of stimulus current. All stimulating sites of the MoV were located inside the MoV and the shortesl distance between stimulating sites and the boundary of MoV was larger than 160/~m in 15 AT neurons (mean 274 + 104.6 /~m). We used stimuli of intensity less than 2(} # A , therefore the possibility of activating fibers near the MoV should be low. We applied the spike triggered averaging method (STA)40,¢~.~ to 4 rhythmically activated AT neurons to obtain further confirmation of the projection of these AT neurons to the masseteric m o t o n e u r o n pool. For all these neurons, a negative or positive field potential was induced in the MaMP. In 3 of these neurons brief biphasic or triphasic potentials, preceding the positive or negative field potential, appeared with latencies of 0.19-0.43 ms after the onset of spike potentials of the trigger neuron. The latency of these brief potentials was shorter than that of the antidromic spike potentials for each trigger neuron by 0.17-0.25 ms when the MoV was stimulated. If this difference in the latency is attributed to the utilization time for initiation of the spike potential of the trigger neuron ~7 the brief potential would be regarded as the presynaptic terminal potential of the trigger neuron. The positive or negative field potentials followed 0.42-0.68 ms after the onset of the bi- or triphasic brief potentials, and such latency was estimated as the synaptic delay ~s. Therefore, it is suggested that those positive or negative field potentials reflected the postsynaptic potentials (PSPs) of the masticatory motoneurons evoked by activation of the trigger neurons. Consequently, these 4 neurons were most likely trigeminal premotor neurons. However, we could not identify them to be excitatory or inhibitory neurons because the polarity may change according to the position to the MoV ¢,3

Rhythmic actil~ities of neurons in the SupV and RfmVo The neurons which showed rhythmical burst discharge during rhythmical masticatory activities were reported to be located in the SupV of the chronic rat ~6 and in the lateral ponto-medullary reticular formation of the anesthetized rat 3~. Olsson et al, 49 reported a neuron which fired a burst of spikes during jaw closure and was located in Regio h which is just posterior to the MoV in the anesthetized cat. The discharge characteristics of this neuron resembled that of the present type 1 neurons which fired during the jaw-closing phase of CIFM. In above studies, however, the rhythmic activities of the reticular neurons could have resulted from activity in phasic peripheral inputs since those animals were not immobilized. In the present study, we found RA neurons in the SupV and the RfmVo. Since the animals were immobilized, rhythmic activities of the RA neurons were produced centrally by a rhythm generator, as opposed to phasic activity in peripheral receptors. Recently, Donga and Lund 7 recorded from rhythmically active neurons in the intertrigeminal area and nucleus oralis 3' of the immobilized rabbit which were antidromically activated from the contralateral trigeminal motor nucleus. On the other hand, Kubo et al. 25 reported that the SupV neurons of the immobilized cat did not show rhythmic activities during repetitive cortical stimulation. The difference between their study and ours may partly be due to species difference and partly to different sampling of neurons. For example, the cat's SupV neurons responded with the

289 burst of multiple spikes to stimulation of the trigeminal nerve (lingual nerve) at short latencies (mean 1.88 + 0.37 ms) 23 and did not respond to single pulse or short train pulse stimulation of cortex 25, while most of the rat's RA neurons in SupV recorded in the present study responded with few spikes at longer latencies (mean 5.8 + 2.41 ms) or were sometimes inhibited to trigeminal stimulation (IA), and 40% of the rat's RA neurons in SupV responded to triple pulse stimulation of cortex. Peripheral and cortical input to RA neurons Thirty-nine (91%) of the 43 RA neurons and 8 (89%) neurons of 9 A-RA neurons responded to at least one of the tested peripheral stimuli and most of them received peripheral convergence. Jerge 19 showed all of units in the SupV responded to peripheral inputs and many SupV units received peripheral convergence in the anesthetized cat. Donga et al. 8 reported that many trigeminal commissural last-order interneurons in the intertrigeminal area, nucleus oralis 3/ and SupV received convergent inputs from both mandibular and maxillary divisions in the anesthetized rabbit. Westberg and Olsson 65 also showed that interneurons in the nucleus oralis 3, received peripheral convergent inputs in the anesthetized cat. The results of the above reports are in accord with those of the present study. It also should be noted that peripheral sensory inputs are not always excitatory but sometimes inhibitory to these reticular neurons, e.g., 8 (38%) of the 21 SupV neurons and one (9%) of the 11 RfmVo neurons showed inhibitory responses to at least one of the tested peripheral stimuli. In the present study, RA neurons in the RfmVo responded to IA stimulation at short latencies of 1.8-6.1 ms (mean 3.1 + 1.17 ms). This is in good agreement with the latencies of 1.2-26.4 ms (median 3.0 ms) of interneuronal discharge in the nucleus oralis 3' in the cat evoked by IA stimulation 65. On the other hand, RA neurons in the SupV responded at longer latencies of 3.4-9.6 ms (mean 5.8 ___2.25 ms) to IA stimulation than the RfmVo neurons. Sensation from IA seems to activate RA neurons in the RfmVo more directly than those in the SupV. We examined response to triple-pulse stimulation of the contralateral cortex with 31 RA neurons and 15 (48%) were found to respond at a latency of 8.3-36.8 ms (mean 17.5 + 9.9 ms). On the other hand, only 5 (26%) of the 19 RA neurons responded to the ipsilateral cortical stimulation at a latency of 9.9-48.0 ms (mean 22.2 + 12.8 ms). Most of those RA neurons responding to cortical stimulation were located in the SupV and the RfmVo. From their long latencies, they were supposed to receive cortical inputs polysynaptically. Comparison of responses between the reticular neurons around the MoV and those in the medial bulbar area Premotor neurons that were involved in the cortically evoked inhibition of the jaw-closing motoneurons or excitation of the jaw-opening motoneurons, are found in the medial bulbar reticular formation in the cat 12,21,46,61. Furthermore, Hiraba et al. ~2 showed in the chronic cat some neurons in this area were rhythmically activated during mastication and projected to the MoV. Based on these results, they concluded that these neurons may participate in the central control of rhythmical ingestive movements of the jaw. There are differences between the neurons of the present study and those found in the medial bulbar reticular formation. First, regarding the phase of activation of the neurons during fictive or actual mastication, type 1 neurons (44%) were the most numerous of all types in the present study and the phase in which those neurons were activated might correspond to the jaw-closing phase during mastication. On the other

29(t hand, only 15% of the medial bulbar reticular neurons showed a spike burst or attained the maximum firing frequency during the jaw-closing phase 12 Secondly, as to peripheral inputs, the RA neurons in the present study received abundant peripheral inputs, while a few premotor neurons in the medial bulbar reticular formation responded to stimulation of the lingual nerve after a long latency 46. Thirdly, we did not observe monosynaptic cortical inputs in the present study though the medial bulbar reticular neurons were shown to receive monosynaptic cortical inputs 4~, The above differences suggest different functional roles for the neurons around the MoV and medial bulbar reticular formation. The neurons of the present study receive peripheral inputs and may be concerned with sensory feedback control of the masticatory muscles, while the medial bulbar reticular neurons may be involved in the cortically evoked inhibition and excitation of the masticatory motoneurons a n d / o r in the central generation of the rhythmical masticatory activity (,.12 Functional roles of the RA neurons Type 1 RA neurons were activated during the phase of masseteric activity. Food is actually broken or ground during this phase, and activities of the jaw-closing muscles during this phase are controlled according to physical properties of food. Inputs from intraoral structures and muscle spindles play important roles in control of closer muscle activity. It was reported that inputs from the periodontal and muscle spindle afferents enhanced jaw-closing muscle activities through a positive feedback loop 2s,3s and stimulation of IA 23 evoked inhibition of jaw-closer motoneurons. In the present study, type 1 neurons responded to IA, IO and passive jaw opening, and some of them showed inhibitory responses to peripheral nerve stimulation. These findings suggest that type 1 neurons receive inputs fiom the muscle spindles, temporomandibular joint, intraa n d / o r perioral structures. Three of the 9 antidromically activated RA neurons were classified as type 1 neurons and are likely to be trigeminal premotor neurons which receive phasic excitatory input from the CPG during jaw closing and mediate reflex excitation of jaw-closer motoneurons during muscle spindle or intraoral stimulation. Two of the 3 AT neurons were located in the SupV. Recently it was shown that glutamatergic trigeminal premotor neurons were located in the SupV (Chandler, personal communication), suggesting the presence of excitatory premotor neuron in this region. This peripheral and central convergence onto common last-order interneurons provides for appropriate level of modulation of closer motoneuron activity during mastication. A similar discussion is applicable to type 3 neurons which were activated during the phase of digastric activity. They may be either excitatory premotor neurons for digastric motoneurons or the inhibitory premotor neurons for masseteric motoneuron. Type 2 and type 4 neurons were activated during the transition phase between masseteric and digastric activity. We recorded some of these neurons in the bulbar parvocellular reticular formation, previously shown to contain facial and hypoglossal premotor n e u r o n s 36,51,59,60,62,64. Recently, Amri et al. 2 demonstrated that some neurons in the ventral reticular formation of the medulla have branched axons innervating both trigeminal and hypoglossal motor nuclei. That study did not examine the regions where neurons were recorded in the present study. It is likely that type 2 and type 4 neurons contribute to coordination of the masticatory, facial and tongue muscles. In addition to the above mentioned functions of trigeminal premotor neurons, some of the RA neurons may modulate activities of masticatory motoneurons indirectly by affecting activities of premotor neurons which transmit sensory information from the

291 periphery to these motoneurons. Lund and Olsson 3a demonstrated that the amplitude of the jaw-opening reflex was modulated in a phase dependent manner during the cortically induced rhythmic jaw movements, and was confirmed later by Morimoto et al. 37. It is likely that this modulation resulted from a change in the excitability of the second-order neurons, since jaw opener motoneurons were not hyperpolarized during mastication H,42. Olsson et al. 49 found that some of the second-order neurons were depressed during the jaw closing and occlusion phases of the rhythmic jaw movements in the anesthetized rabbit. It is also possible that some of the RA neurons may modulate presynaptically the afferent inputs from the muscle spindles and periodontal afferents to the jaw-closing motoneurons through the mesencephalic nucleus. Periodontal and muscle afferents originating from the trigeminal mesencephalic nucleus often terminate in the MoV directly 9,45,53,54,56. In our previous study 38, these mesencephalic inputs to the jaw-closing motoneurons were generally facilitatory. Accordingly, presynaptic changes in activities of these afferents would result in the change in activities of the jaw-closing motoneurons. Another possible function of type 2 and 4 neurons is that some of them may play a role in altering the timing of activities of the masseteric or digastric muscles, since they were activated in the transition phase between masseteric and digastric activity. Though reticular neurons around the MoV are not critical for the production of rhythmical jaw movement 6,47, type 2 and 4 neurons might adjust the timing of activation of masticatory muscle by affecting phasic transmission from a central timing network for mastication to trigeminal motoneurons. As discussed above, we propose that the RA neurons around the MoV may function as (1) excitatory or inhibitory premotor neurons projecting to the MoV, facial nucleus and hypoglossal nucleus which transmit inputs from the central and/or peripheral structures, (2) the interneurons modulating activities of the above premotor neurons or those modulating presynaptically activities of monosynaptic inputs of the muscle spindle or periodontal afferents through the mesencephalic trigeminal nucleus to the MoV. ACKNOWLEDGEMENTS

We thank Dr. S.H. Chandler for reading and editing the manuscript. This study was supported by grants-in-aid (Nos. 01771506 and 01304045) from the Japanese Ministry of Education, Science and Culture. REFERENCES 1 Abzug, C. Maeda, N. Peterson, B.W. and Wilson, V.J., Cervical branching of lumbar vestibulospinal axons, J. PhysioL 243 (1974) 499-522. 2 Amri, M. Car, A. and Roman, C., Axonal branching of medullary swallowing neurons projecting on the trigeminal and hypoglossal motor nuclei: demonstration by electrophysiological and fluorescent double labeling techniques, Exp. Brain. Res., 81 (1990) 384-390. 3 Appenteng, K. Conyers, L. and Moore, J.A., The monosynaptic excitatory connections of single trigeminal interneurones to the V motor nucleus of the rat, J. PhysioL 417 (1989) 91-104. 4 Appenteng, K. Conyers, L. Curtis, J. and Moore, J.A., Monosynaptic connexions of single interneurones to the contralateral V motor nucleus in anaesthetised rats, Brain Res., 514 (1990) 128-130. 5 Asanuma, H., Microstimulation technique, In Patterson, M.M. and Kesner R.P. (Eds.), Electrical stimulation Technique, Academic Press, New York, NY, 1981, pp. 61-70. 6 Chandler, S.H. Turman Jr., J. Salem, L. and Goldberg, L.J., The effects of nanoliter ejections of lidocaine into the pontomedullary reticular formation on cortically induced rhythmical jaw movements in the guinea pig, Brain Res., 526 (1990) 54-64.

292 7 Donga, R. and Lund, J.P., Discharge patterns of identified trigemimd interneuroncs during I:ict6c mastication in the anaesthetized rabbit, jr. PhysioL, 423, 74P, 199(). 8 Donga, R. Lnnd, J.P. Veilleux, D, An electrophysiological study of trigeminal commissural interncurons m the anaesthetized rabbit, Brain Rex., 515 (19901 351-354. 9 Dessem, D. and Taylor, A., Morphology of jaw-muscle spindle afferents in the rat, J. Comp. Neurol., 2,~2 (1989) 389-403. 10 Goldberg, L.J. and Nakamura, Y., Lingually induced inhibition of masseteric motoneurons, Experientm. 24 (19681 371-373. 11 Goldberg, L.J. Chandler, S.H. and Tal, M. Relationship between jaw movements and trigeminat motoneuron membrane-potential fluctuatkms during cortically induced rhythmical jaw movements in the guinea pig, J. Neurophysiol. 58 (1982) 102-125. 12 Hiraba, K. Taira, M. Sahara, Y. and Nakamura. Y., Single-unit activity in bulbar reticular formation during food ingestion in chronic cats, J. Neurophysh~l., 6(I (19881 1333-1349. 13 Hopkins, D.A. and Holstege, G., Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat, Exp. Brain. Res., 32 (1978) 529-547. 14 Inoue, T. Kato, T. Masuda, Y. Nakamura, T. Kawamura, Y. and Morimoto, T., Modification of masticatory behavior after trigeminal deafferentation in the rabbit, Exp. Brain. Res., 74 (1989) 579-591. 15 lnoue, T. Masuda, Y. Saito, O. Nagashima, T. and Morimoto, T., Convergence of cortical and peripheral inputs to brain stem neurons in the rat, Jpn..L PhysioL, 39 (1989) 5117. 16 Inoue, T. Masuda, Y. Saito, O. Nagashima, T. and Morimoto, T., Input-output properties of reticular neurons around the trigeminal motor nucleus in the rat, Neurosci. Res., 11 (1990) $68. 17 Jankowska, E. and Roberts, W.J., An electrophysiological demonstration of the axonal projections of single spinal interneurones in the cat, J. PhysioL, 222 (1972) 597-622. 18 Jankowska, E. and Roberts, W.J., Synaptic actions of single interneurones mediating reciprocal la inhibition of motoneurones, J. Physiol., 222 (19721 623-642. 19 Jerge, C.R., The function of the nucleus supratrigeminalis, J. Neurophysiol., 26 (1963) 393-4(12. 20 Kamogawa, H. Hanashima, N. Naito, K. and Kagaya, K., Candidate interneurons mediating peripherally evoked disynaptic inhibition of masseter motoneurons of both sides, Neurosci. Lett., 95 (1988) 149-154. 21 Katoh, M. Taira, M. Katakura, N. and Nakamura, Y., Cortically induced effects on trigeminal motoneurons after transection of the brainstem at the pontobulbar junction in the cat, Neurosci. Lett., 33 (1982) 141-146. 22 Kidokoro, Y. Kubota, K. Shuto, S. and Sumino, R., Reflex organization of the cat masticatory muscles, ./. NeurophysioL, 31 (1968) 695-708. 23 Kidokoro, Y. Kubota, K. Shuto, S. and Sumino, R., Possible interneurons responsible for reflex inhibition of motoneurons of jaw-closing muscles from the inferior dental nerve, J. Neurophysiol., 31 (19681 709-716. 24 Krettek, J.E. and Price, J.L., Amygdaloid projections to subcortical structures within the basal forebrain and brain stem in the rat and cat, .L Comp. Neurol., 178 (19781 225-254. 25 Kubo, Y. Enomoto, S. and Nakamura, Y., Synaptic basis of orbital cortically induced rhythmical masticatory activity of trigeminal motoneurons in immobilized cats, Brain Res., 230 (19811 97-110. 26 Kuypers, H.G.J.M., An anatomical analysis of cortico-bulbar connexions to the ports and lower brain stem in the cat, .LAnat., 92 (1958) 198-218. 27 Landgren, S. Olsson, K.A. and Westberg, K.G., Bulbar neurones with axonal projections to the trigeminal motor nucleus in the cat, Exp. Brain Res., 65 (1986) 98-111. 28 Lavigne, G. Kim, J.S. Valiquette, C. and Lund J.P., Evidence that periodontal presso-receptors provide positive feedback to jaw closing muscles during mastication, J. Neurophysiol., 58 (1987) 342-358. 29 Lorente de No, R., Contribucion al conocimiento del nervio trigemino, Libro en Honor de Ram6n y Cajal, 2 (1922) 13-30. 30 Lorente de No, R,, Vestiblo-ocular reflex arc, Arch. NeuroL Psychiat., 30 (1933) 245-291. 31 Lurid, J.P. Olsson, K.A., The importance of reflexes and their control during jaw movement, Trends Neurosci., 6 (1983) 458-463. 32 Mizuno, N., Projection fibers from the main sensory trigeminal nucleus and the supratrigeminal region, Z Comp. NeuroL, 139 (1970) 457-472. 33 Mizuno, N. Nomura, S. Itoh, K. Nakamura, Y. and Konishi, A., Commissural interneurons for masticatory motoneurons: a light and electron microscope study using the horseradish peroxidase tracer technique, Exp. NeuroL, 60 (1978) 254-262. 34 Mizuno, N. and Sauerland, E.K., Trigeminal proprioceptive projections to the hypoglossal nucleus and the cervical ventral gray column, J. Comp. Neurol., 139 (1970) 215-226. 35 Mizuno, N. Sauerland, E.K. and Clemente, D.C., Projections from the orbital gyrus in the cat. 1. To brain stem structures, J. Comp. NeuroL, 133 (1968) 463-476.

293 36 Mizuno, N. Yasui, S. Nomura, S. Itoh, K. Konishi, A. Takada, M. and Kudo, M., A light and electron microscope study of premotor neurons for the trigeminal motor nucleus, J. Cornp. Neurol., 215 (1983) 290-298. 37 Morimoto, T. Inoue, T. and Kawamura, Y., Modulation of cortical sensory evoked potential and jaw opening reflex during mastication in the rabbit, In Bandler, R. (Ed.), Modulation of Sensorimotor Activity During Alterations in Behavioral States, Alan R. Liss, New York, NY, 1984, pp. 291-295. 38 Morimoto, T. Inoue, T. Masuda, Y. and Nagashima, T., Sensory components facilitating jaw-closing muscle activities in the rabbit, Exp. Brain Res., 76 (1989) 424-440. 39 Moriyama, Y., Rhythmical jaw movements and lateral ponto-medullary reticular neurons in rats, Comp. Biochem. Physiol., 86A (1987) 7-14. 40 Nakamura, Y. Enomoto, S. and Katoh, M., The role of medial bulbar reticular neurons in the orbital cortically induced masticatory rhythm in cats, Brain Res., 202 (1980) 207-212. 41 Nakamura, Y. Goldberg, L.J. Mizuno, N. and Clemente C.D., Effects of hypoglossal afferent stimulation on masseteric motoneurons in cats, Exp. Neurol., 61 (1978) 1-14. 42 Nakamura, Y. Kubo, Y., Masticatory rhythm in intracellular potential of trigeminal motoneurons induced by stimulation of orbirtal cortex and amygdala in cats, Brain Res., 148 (1978) 504-509. 43 Nakamura, Y. Mori, S. and Nagashima, H., Origin and central pathways of crossed inhibitory effects of afferents from the masseteric muscles on the masseteric motoneuron of the cat, Brain Res., 57 (1973) 29-42. 44 Nishimori, T. Sera, M. Suemune, S. Yoshida, A. Tsuru, Y. Tsuiki, Y. Akisaka, T. Okamoto, T. Dateoka, Y. and Shigenaga, Y., The distribution of muscle primary afferents from the masseter nerve to the trigeminal sensory nuclei, Brain Res., 372 (1986) 375-381. 45 Nomura, S. and Mizuno, N., Differential distribution of cell bodies and central axons of mesencephalic trigeminal nucleus neurons supplying the jaw-closing muscles and periodontal tissue: A transganglionic tracer study in the cat, Brain Res., 359 (1985) 311-319. 46 Nozaki, S. Enomoto, S. and Nakamura, Y., Identification and input-output properties of bulbar reticular neurons involved in the cerebralcortical control of trigeminal motoneurons in cats, Exp. Brain Res., 49 (1983) 363-372. 47 Nozaki, S. Iriki, A. and Nakamura, Y., Localization of central rhythm generator involved in corticaly induced rhythmical masticatory jawopening movement in the guinea pig, J. Neurophysiol., 55 (1986) 806-825. 48 Ohta, M. Moriyama, Y., Supratrigeminal neurons mediate the shortest, disynaptic pathway from the central amygdaloid nucleus to the contralateral trigeminal motoneurons in the rat, Comp. Biochem. PhysioL, 83A (1986) 633-641. 49 Olsson, K.A. Sasamoto, K. and Lund, J.P., Modulation of transmission in rostral trigeminal sensory nuclei during chewing, J. Neurophysiol., 55 (1986) 56-75. 50 Olsson, K.,~. and Westberg, K.-G., Integration in trigeminal premotor interneurones in the cat. 2. Functional characteristics of neurones in the subnucleus-y of the oral nucleus of the spinal trigeminal tract with a projection to the digastric motoneurone subnucleus, Exp. Brain Res., 84 (1991) 115-124. 5l Porter, R., Cortical actions on hypoglossal motoneurones in cats: A proposed role for a common internuncial cell, J. Physiol. 193 (1967) 295-308. 52 Sasamoto, K. Zhang, G. and Iwasaki, M., Two types of rhythmical jaw movements evoked by stimulation of the rat cortex, Jpn. Z Oral Biol. , 32 (1990) 57-68. 53 Shigenaga, Y. Doe, K. Suemune, S. Mitsuhiro, Y. Tsuru, K. Otani, K. Shirana, Y. Hoshi, M. Yoshida, A. and Kagawa, K., Physiological and morphological characteristics of periodontal mesencephalic trigeminal neurons in the cat - intra-axonal staining with HRP, Brain Res., 505 (1989) 91-110. 54 Shigenaga, Y. Mitsuhiro, Y. Yoshida, A. Cao, C.Q. and Tsuru, H., Morphology of single mesencephalic trigeminal neurons innervating masseter muscle of the cat, Brain Res., 445 (1988) 392-399. 55 Shigenaga, Y. Sera, M. Nishimori, T. Suemune, S. Nishimura, M. Yoshida, A. and Tsuru, K., The central projection of Masticatory afferent fibers to the trigeminal sensory nuclear complex and upper cervical spinal cord, J. Comp. Neurol., 268 (1988) 489-507. 56 Shigenaga, Y. Yoshida, A. Mitsuhiro, Y. Doe, K. and Suemune, S., Morphology of single mesencephalic trigeminal neurons innervating periodontal ligament of the cat, Brain Res., 448 (1988) 331-338. 57 Shigenaga, Y. Yoshida, A. Mitsuhiro, Y. Tsuru, K. and Doe, K., Morphological and functional properties of trigeminal nucleus oralis neurons projecting to the trigeminal motor nucleus of the cat, Brain Res., 461 (1988) 143-149. 58 Szenthagothai, J. and Rajkovits, K., Der Hirnnervenanteil der Pyramidenbahn und der pramotorishe Apparat motorischer Hirnnervenkerne, Arch. Psychiat. Z. Ges. Neurol., 197 (1958) 335-354.

294 59 Takada, M. ltoh, K. Yasui, Y. Mitani, A. Nomura, S. and Mizuno, N., Distribution ol premotor ncur~ls for orbicutaris oculi motoneurons in the cat, with particular reference to possible pathways for blink refit:x, Neurosci. Lett., 50 (1984) 25t-255. 60 Takada, M. Itoh, K. Yasui, Y. Mitani, A. Nomura, S. and Mizuno, N., Distribution of premotor neurorls for the hypoglossal nucleus in the cat, Neurosci. Lett., 52 (1984) 141-146. 61 Takatori, N. Nozaki, S. and Nakamura, Y., Control of trigeminal motoneurons exerted from bulb;it reticular formation in the cat, Exp. Neurot, 72 (198l) 122-140. 62 Takeuchi, Y. Nakano, K. Uemura, M. Matsushima, R. and Mizuno, N., Mesencephlic and pontine afferent fiber system to the facial nucleus in the cat: a study using the horseradish peroxidase and silver impregnation techniques, Exp. Neurol., 66 (1979) 330-342. 63 Taylor, A. Stephens, J.A. Somjen, G. Appenteng, K. and O'donovan, M.J., Extracellular spike triggered averaging for plotting synaptic projections, Brain Res., 140 (1978) 344-348. 64 Travers, J.B. and Norgren, R., Afferent projections to the oral motor nuclei in the rat, .I. Comp. Neurol., 220 (1983) 280-298. 65 Westberg, K.-G. and Olsson, K.A., Integration in trigeminal premotor interneurones in the cat. I. Functional characteristics of neurones in the subnucleus-y of the oral nucleus of the spinal trigeminal tract, Exp. Brain Res., 84 (1991) 102-114. 66 Yamamoto, T. Matsuo, R. Kiyomitsu, Y. and Kitamura, R., Sensory and motor responses of trigeminal and reticular neurons during ingestive behavior in rats, Exp. Brain Res., 76 (1989) 386-400. 67 Yasui, Y. Itoh, K. Mitani, A. Takada, M. and Mizuno, N., Cerebral cortical projections to the reticular regions around the trigeminal motor nucleus in the cat, J. Comp. Neurot, 241 (1985) 348-356.