Exp Brain Res (1994) 100:7-17

9 Springer-Verlag 1994

Satoshi Matsuo 9 Masae Hosogai 9 Shozo Nakao

Ascending projections of posterior canal-activated excitatory and inhibitory secondary vestibular neurons to the mesodiencephalon in cats

Received: 22 October 1993 / Accepted: 22 February 1994 Abstract The axonal projections of 62 posterior canal and that PC-related inhibitory neurons seem to take (PC)-activated excitatory and inhibitory secondary part only in the genesis of vertical eye movements. vestibular neurons were studied electrophysiologically in cats. PC-related neurons were identified by monosy- Key words Posterior semicircular canal naptic activation elicited by electrical stimulation of the Vestibular nucleus neuron vestibular nerve and activation following nose-up rota- Medial mesodiencephalic junction - Thalamus 9 Cat tion of the animal's head. Single excitatory and inhibitory neurons were identified by antidromic activation following electrical stimulation of the contralateral and Introduction ipsilateral medial longitudinal fasciculus, respectively. The oculomotor projections of identified neurons were The medial mesodiencephalic junction is an important confirmed with a spike-triggered averaging technique. premotor area in generating vertical eye movements The axonal projections of the identified neurons were (Biittner et al. 1977; Bfittner-Ennever et al. 1978; K6mpf then studied by systematic, antidromic stimulation of et al. 1979; Nakao and Shiraishi 1985). Neurons in the the mesodiencephalon. Excitatory neurons showed two junction including the central gray, interstitial nucleus main types of axonal projections. In one type, axonal of Cajal (INC), and Forel's field H (FFH) are classified branches were issued to the interstitial nucleus of Cajal, into several types by their firing patterns in relation to central gray, and thalamus including the ventral pos- spontaneous vertical eye movements and visual- or terolateral, ventral posteromedial, ventral lateral, ven- vestibular-induced vertical eye movements (B/ittner et tral medial, centromedian, central lateral, lateral poste- al. 1977; King and Fuchs 1979). Anatomical study in rior, and ventral lateral geniculate nuclei. The other monkeys has shown that vestibular neurons send axons type was more frequently observed, giving off axon col- to the INC and rostral interstitial nucleus of medial lonlaterals to the above-mentioned regions and to Forel's gitudinal fasciculus (riMLF) (Bfittner-Ennever et al. field H as well. Inhibitory neurons issued axonal 1978). The riMLF in monkeys corresponds to the medibranches to limited areas which included the central al part of the F F H in cats (Graybiel 1977). To clarify the gray, interstitial nucleus of Cajal, its adjacent reticular composition of a three-dimensional eye movement, conformation and caudalmost part of Forel's field H, but sideration was given to the contribution of single cells to not the rostral part of the Forel's field H and the thala- vertical or torsional eye movemenfs, and also to less mus. These results suggest that PC-related excitatory efficacious canal-muscle relationships other than the neurons participate in the genesis of vertical eye move- principal canal-muscle relationships. ments and in the perception of the vestibular sensation, Graf and colleagues (1983, 1986) studied axonal projections of single vertical canal-related secondary vestibular neurons in cats and rabbits by means of intraS. Matsuo (l~) axonal injection of horseradish peroxidase. Their studDepartment of Otorhinolaryngology, Faculty of Medicine, ies showed that posterior canal (PC)-related neurons Tottori University, 86 Nishi-machi, Yonago 683, Japan, give off axon collaterals that ascend to the extraocular FAX no: 81-859-34-8090 motoneuron pools. The collaterals extended as far rosM. Hosogai 9 S. Nakao 1 trally as the INC and prerubral field. However, little Department of Physiology, Faculty of Medicine, informations are available on the projection pattern of Tottori University, 86 Nishi-machi, Yonago 683, Japan the single neurons in the premotor areas, particularly in 1Deceased the FFH, and also on the projection of axons to other

regions more rostral to the premotor areas. The vestibular nuclei are known to issue a minor projection to the thalamus (Sans et al. 1970; Lang et al. 1979; Kotchabhakdi et al. 1980; Mergner et al. 1981; Nagata 1986; Isu et al. 1991). The vestibulothalamic neurons convey vestibular information to the cortex via thalamic neurons (Sans et al. 1970; Liedgren and Schwartz 1976; Mergner et al. 1981). Presumably, single PC-related vestibular neurons project to the thalamus, and subclassification of the neurons may depend on the difference in the pattern of projections. However, it is uncertain whether these vestibulothalamic neurons also project to the oculomotor nucleus or to other premotor areas, and whether synaptic action of the neurons is excitatory or inhibitory. The present study was undertaken to investigate the branching pattern of axons of single PC-related excitatory and inhibitory secondary vestibular neurons in the mesodiencephalon, with special attention directed to projections to the oculomotor nucleus, medial mesodiencephalic junction, and thalamus, using the antidromic microstimulation technique. A part of this study has been published in a preliminary report (Matsuo et al. 1992).

Materials and methods Preparations Experiments were performed on 30 cats weighing 2.5-4.2 kg. All procedures used in this study were reviewed and fulfilled the requirements of "Guidelines for animal experimentation" of the Faculty of Medicine, Tottori University and conformed to standards set forth in the National Institutes of Health's "Guide for the care and use of animals." The animals were anesthetized with sodium pentobarbital (Nembutal: 40 mg/kg, with additional doses when necessary). The state of anesthesia was monitored by observing blood pressure, heart rate and pupillary size throughout the experiment (Ohgaki.et al. 1989). The animals were considered to be in a pain-free state when the heart rate was regular and less than 180/rain, when mean arterial pressure monitored from a femoral artery was stable between 100 and 160 mmHg, and when the pupil was either slit-like or in mid-dilatation. The cats were mounted in a stereotaxic apparatus which could be sinusoidally rotated. The pitch of the apparatus could be controlled by hand. PC-related excitatory neurons project to the inferior rectus (IR) subdivision of the oculomotor nucleus through the contralateral MLF (Uchino et al. 1981; Uchino et al. 1982). PC-related inhibitory neurons project to the superior rectus (SR) subdivision through the ipsilateral MLF (Uchino et al. 198l; Uchino and Suzuki 1983). On the basis of these well-known facts, experiments were carried out using the following two groups: (1) In 17 animals, a pair of Ag-AgC1 electrodes were implanted on the left round and oval windows with intact receptors by the ventral approach to stimulate the vestibular nerve. The right IR nerve was detached from the IR muscles and mounted on an Ag-AgC1 hook bipolar electrode for antidromic stimulation. The right parieto-occipital lobe was aspirated to expose the underlying superior colliculus and dorsal surface of the thalamus, and the medial cerebellum was also aspirated to visualize the floor of the fourth ventricle. Three stimulating electrodes were placed in the right MLF, IR subdivision, and mesodiencephalon to test the projection of PC-related vestibular neurons. The first stimulating electrode (acupuncture needle electrode; 200/am diameter, about 0.4 M ~ resistance) was placed in the right MLF. The final position of the electrode was determined by monitoring the characteristic orthodromic field

potentials evoked from vestibular nerve stimulation. The second stimulating electrode (acupuncture needle microelectrode; 200/am diameter, about 0.4 M ~ resistance) was inserted into the right IR subdivision through the exposed superior colliculus, where the antidromic field potential evoked by the IR nerve stimulation was maximally recorded (2-3 mV). The third systematic stimulating electrode (a glass-insulated steel electrode; 30/am in tip diameter, about 1.0 Mf~ resistance) was then introduced in the right mesodiencephalon (Fig. 1A); (2) In another 13 cats, Ag-AgC1 electrodes were implanted on the right round and oval windows. The left SR nerve was detached from the SR muscles and mounted on an Ag-AgC1 hook electrode. To test the projection of PC-related neurons to the right MLF, SR subdivision and mesodiencephalon, three stimulating electrodes were placed in the respective regions (Fig. 6A). During the recording period, the animal was immobilized by intravenous administration of pancuronium bromide, 0.4 mg/kg initially and 0.2-0.4 mg subsequently (Mioblock, Organon) and artificially ventilated with room air. Bilateral pneumothorax was performed to minimize movement of the brain stem. Rectal temperature was maintained at 36-38~ by a heating pad. Stimulation Rectangular cathodal currents of 0.1 ms duration were passed through one of the stimulating electrodes. Stimulus currents monitored by the voltage drop across a 100-~ resistor were usually restricted to less than 30/aA. Recording Glass micropipettes filled with 2 M NaCI solution (about 1.0 Mf~ resistance) saturated with Fast green FCF dye (Thomas and Wilson 1965) were used for recording extracellular unit spikes. Activities of single PC-related excitatory neurons were recorded from the left vestibular nucleus (Fig. 1A), and those of the inhibitory neurons from the right vestibular nucleus (Fig. 6A). The potentials were fed to an oscilloscope (VC-11, Nihon Kohden) via a DC amplifier and were recorded on film. Spike-triggered averaging To study synaptic action of PC-related neurons in the IR or SR subdivision, in most cases, the activity of each subdivision was recorded with the stimulating microelectrode placed in each region and was fed into a DC amplifier. Spike potentials of the single PC-related neuron were used to trigger an averaging computer (QC-111J, Nihon Kohden). Compound action currents of the subdivision were averaged over a 10-ms sweep duration after, or before and after, triggering spikes. Electrophysiological study of axonal projection Axonal projections of single identified PC-related neurons to the mesodiencephalon were examined using the systematic, antidromic mapping technique (Jankowska and Roberts 1972). The systematic stimulating electrode was advanced from the dorsal surface of the thalamus to deeper structures. Then, for each track, threshold currents and latencies for antidromic activation of single PC-related secondary vestibular neurons were measured at 10G/am intervals. Histological procedure At the end of each experiment we made small, electrolytic lesions and Fast green FCF dye spots at the stimulating sites in the brainstem and recording sites of the PC-related neurons (Thomas and

9

shocks to the contralateral MLF and the IR subdivision (Fig. 1E, F). Activation of the neuron from the contralateral MLF and the IR subdivision seemed to be antidromic. Its antidromic nature was further confirmed by a collision block of the stimulUs evoked spikes with preceding vestibular nerve-evoked spikes (Fig. 1G-I). All the excitatory neurons were verified to be located in the border region between the medial and descending Results vestibular nuclei. Synaptic potentials evoked in the IR subdivision Identification of PC-activated excitatory secondary were studied in 25 of the 36 excitatory neurons with the vestibular neurons spike-triggered averaging technique (Fig. 2A). As shown We identified 36 PC-activated excitatory secondary by an antidromic activation (Fig. 2B), the PC-related vestibular neurons. Figure 1B-I illustrates spike re- vestibular neurons projected to the contralateral IR sponses of an identified neuron. In response to vertical subdivision. With the same stimulating microelectrode, head rotation in pitch, the neuron discharged spikes simultaneous recording was made of spikes of a PC-reduring a nose-up rotation, but not during nose-down lated neuron and the extracellular field potential in the head rotation (Fig. 1B). The neuron was nearly silent IR subdivision. The field potential was averaged with when the head was still (Fig. 1C). Stimulation of the ipsi- the rotation-induced spikes of the neuron as the trigger lateral vestibular nerve induced spikes of the neuron on (Fig. 2Ca): the averaged field potential revealed a posithe N1 field potentials (Shimazu and Precht 1965) tive-negative spike followed by a negative wave (Fig. 1D). The PC-related neuron further discharged (Fig. 2Cb). The succeeding negative wave had a spikes with a short and fixed latency following single monosynaptic latency of 0.9 ms when measured between the onset of the triggering spike (vertical broken line in Fig. 2C) and wave onset. The negative wave had Fig. 1A-I Identification of a single PC-related excitatory neuron. a steep falling phase and a relatively long recovery A Schematic diagram of the experimental arrangement. B Spike phase. All the 25 PC-related excitatory neurons induced responses of the excitatory neuron (a) during pitch rotation of the animal's head (b). C The neuron is nearly silent (a), when the head the characteristic spike and negative wave complexes in is still (b). Upward deviation of trace b indicates head movements the averaged field potentials recorded from the IR subin the nose-up direction. D Monosynaptic activation of the neu- division. Latencies of the negative wave from the trigron following stimulation of the ipsilateral labyrinth. The spikes gering spikes were 0.5-1.3 ms (mean 0.7 ms, n = 25). Wilson 1965), respectively, by passing cathodal currents through the electrodes for later reconstruction. The animal was then sacrificed by intravenous administration of a lethal dose of pentobarbital, and the brain was removed and fixed in a 10% formalin solution. The recording and stimulating sites were histologically verified in 50flxm-thick, Nissl-stained serial sections of the brainstem.

are evoked on the N1 field potentials with a short latency (1.0 ms). E,F Spike responses of the neuron to the single shocks to the contralateral M L F at threshold intensity (2.0 gA) and to the IR subdivision at threshold intensity (1.3 IxA). G--I Collision-block testing. IR subdivision-induced antidromic spikes of the same neuron are present (G), but are blocked by preceding labyrinthinduced spikes (H) at short intershock intervals. In I, labyrinthinduced spikes are absent and only N1 field potentials are present. Upward arrows denote the occurrence of stimuli. (llI oculomotor nucleus, L labyrinth, 7/t thalamus, VN vestibular nucleus)

A

Electrophysiological studies of the axonal projection of single PC-related excitatory secondary vestibular neurons The identified single excitatory neurons were subjected to systematic threshold mapping by evoking antidromic

B TH =F M . ,,.:-:

.J-L / s,~'tematic ~,,--;trscking

a

['--"

b -

_ J-L

~'~

nose-up nose-down

:

I

111I II II!1 I! . . . . . . . . . . . . . . . . . . . . . . .

, : :[

II

200

D

E

mV

F

G

H

I

ms

$

10 Fig. 2A-C Averaged negative field potentials evoked in the IR subdivision by spike-triggered averaging of a PC-related excitatory neuron. A Schematic diagram of the experimental arrangement. B Antidromic responses of a PC-related excitatory neuron evoked by single (a) and double shocks (b) delivered to the field potential recording site. Upward arrows denote the occurrence of stimuli. C Averaged field potentials (1500 sweeps) in the IR subdivision (b) triggered by spikes of the neuron (a); vertical broken line indicates foot of triggering spike. (III oculomotor nucleus, VN vestibular nucleus)

A

B _i-L _o

! ~'~ ; ; " z % ', ~ ', ~ " , i ('"

I

9 trigger

mV

b

s "" :
--

,, ,-', !,, 1 ms

A i :

i~ :

;

i ! :

:

:

PCi~i

'i! :'i~

i {! {

. -

Ii,i<,,

MDI:

!H'-'

i ~ 9 . s

i i\j!i

i

~:iiil

',i

i :

})!

f-i i |

B

ct, t

f

:

ii/

i

1.5

i

:

~

! 1:3!

~

~

:

~

VP]

i ! i/i-i-~"

~IS" i '1ram'

O <2pA O 2-10 pA o 10-20 pA o 20-30 pA 30 IJAC

b Fig. 3 A Axonal projections of a PC-related excitatory neuron to the upper brain stem on A 5.0, A 5.5, A 6.0, A 6.5 and A 8.0 Horsely-Clarke planes. The size of each circle indicates the threshold current for antidromic activation of the neuron as calibrated in the right bottom 9 Numerals by the effective sites indicate antidromic latencies in ms. B Antidromic spikes evoked by the stimulation of the contralateral M L F (a), IR subdivision (b), cen-

tral gray (c), INC (d) and thalamus (e). f, g Collision-block testing. Thalamus-induced spikes are blocked by the preceding IR subdivision-induced spikes. Upward arrows denote the occurrence of stimuli9(CL central lateral nucleus, MD medial dorsal nucleus, PC posterior commissure, 3N oculomotor nerve, VPM ventral posteromedial nucleus)

tl

"'L//

/ ::

"

."---"-

-

/

/ CG/

~3.4

'3.4

PL

a.8

/ 0 10-20/=,6," o 20-30 9

Fig. 4 Axonal projections of a PC-related excitatory neuron to the FFH and thalamus on A 7.0 and A 8.0 Horsely-Clarke planes. (CG central gray, C M centromedian nucleus, F R fasciculus retroflexus, L G lateral geniculate nucleus, L P lateral posterior

~A

3 0 ~A<

nucleus, M D medial dorsal nucleus, M G medial geniculate nucleus, VPL ventral posterolateral nucleus, VPM ventral posteromedial nucleus)

J

A 8.5//'''~

A 6.5

_

N

spikes in the m e s o d i e n c e p h a l o n : the m a j o r i t y of neurons projected to the m e s o d i e n c e p h a l i c j u n c t i o n , which includes the central gray, I N C , F F H , a n d thalamus. Figure 3 shows a schematic m a p of the distribution of effective sites for a n t i d r o m i c activation of a single excitatory n e u r o n . T h e size of circles a n d the n u m e r a l s by the circles indicate the levels of threshold a n d latency, respectively. T h e effective sites were in the I N C a n d

Fig. 5 Distribution of effective sites in the mesodiencephalon for antidromic activation of 18 PC-related excitatory neurons. Antidromic spikes were elicited in the hatched areas with a stimulus intensity less than 30 gA. The results of 18 excitatory neurons are superimposed on transverse planes (A 6.5, A 7.5 and A 8.5 Horsely-Clarke planes). (CG central gray, C M centromedian nucleus, F R fasciculus retroflexus, M D medial dorsal nucleus, M G medial geniculate nucleus, M M medial mamillary nucleus, 3N oculomotor nerve, P C posterior commissure, R red nucleus, VPL ventral posterolateral nucleus, VPM ventral posteromedial nucleus)

12

ventral part of the central gray, and the mediocaudal thalamus. Low threshold foci were present between the INC and central gray. The latencies were prolonged in the lateral part of the INC and thalamus. The INC and ventral part of the central gray showed wide variations in thresholds. Figure 4 is a diagram showing the distribution of effective sites for antidromic activation in a representative excitatory neuron. The effective sites are reconstructed on two transverse planes through caudal and rostral levels of the FFH. On the caudal plane (A 7.0 Horsely-Clarke plane), the effective sites were located in the medial part of the F F H and more medially in the ventral part of the central gray. The effective sites on the rostral plane (A 8.0 Horsely-Clarke plane) were also located in the lateral part of the F F H and zona incerta. In this tracking experiment, antidromic spike latencies varied widely among the effective sites in the medial part of the FFH. The latencies were shorter in the medial part of the F F H and central gray than in the lateral part of the FFH. For the same PC neuron, effective sites were also found more dorsolaterally in the ventrocaudal part of the thalamus including the centromedian, central lateral, ventral posteromedial, ventral part of the lateral posterior, ventral posterolateral, and ventral lateral geniculate nuclei. The antidromic latencies were longer in the lateral stimulation sites. Eighteen of the 36 excitatory neurons were antidromically activated from both the F F H and thalamus. These 18 neurons were also antidromically activated from the INC and adjacent mesencephalic reticular formation. The effective sites for the 18 neurons are superimposed on three transverse planes of the brainstem through the levels of the INC, caudal, and rostral parts of the F F H in Fig. 5. The majority of the neurons were antidromically activated from the ventral basal complex

A,, : , ,

Identification of PC-activated inhibitory secondary vestibular neurons We identified 26 inhibitory neurons. Figure 6B-H illustrates the spike responses of an identified inhibitory neuron. In response to pitch rotation, the neuron discharged spikes during nose-up head rotation (Fig. 6B). The neuron spontaneously fired when the head was still (Fig. 6C). When the ipsilateral vestibular nerve was stimulated (Fig. 6D), spikes were evoked on the N1 and N2 field potentials (Shimazu and Precht 1965). The neu-

Fig. 6A-H Identification of a single PC-related inhibitory neuron. A Schematic diagram of the experimental arrangement. B Spike responses of the inhibitory neuron (a) during pitch rotation of the animal's head (b). C The neuron spontaneously fires (a), when the head is stationary (b). Upward deviation of trace b indicates head movements in the nose-up direction. D Monosynaptic activation of the neuron following stimulation of the ipsilateral labyrinth. The spikes are evoked on the N1 and N2 field potentials with a short latency (1.0 ms). Antidromic spikes of the neuron following single shocks at threshold intensity (5.0 t~A) (E) and suprathreshold-double shocks to the ipsilateral M L F (F). The neuron further induces antidromic spikes by single stimulation at threshold currents (4.0 gA) (G) and double stimulation at the suprathreshold currents to the SR subdivision (H). Upward arrows denote the occurrence of stimuli. (III oculomotor nucleus, L labyrinth, TH thalamus, VN vestibular nucleus)

B

9%

_rE

(Y~r /r,-~

systematic tracking

~, ~

of the thalamus, and latencies were longer in the lateral sites. Two neurons were activated from the INC and thalamus, but not from the F F H (Fig. 3). In six other PC neurons, antidromic activation was examined only for the INC, but not for the thalamus. In the remaining ten neurons, antidromic activation was obtained from the IR subdivision, but not from more rostral areas.

)

nose Up nose d o w n

b

--- INC

I ,:2.:;M LF

, l

,.-?--H

"

G

J3

Fig. 7A-C Averaged positive field potentials evoked in the SR subdivision by spike-triggered averaging of a PC-related inhibitory neuron9 A Schematic diagram of the experimental arrangement. B Antidromic responses of a PC-related inhibitory neuron evoked by single (a) and double shocks (b) delivered to the field potential recording site. Upward arrows denote the occurrence of stimuli. C Averaged field potentials (1500 sweeps) in the SR subdivision (b) triggered by spikes of the neuron (a); vertical broken line indicates foot of triggering spike. (III oculomotor nucleus, VATvestibular nucleus)

A

B

__./-,q> ei'oe ~

i~'-__-~-' 8R

a

b 11~

trigger

io. 2mv f

..I;MLF /--~

~ + 1 ms

C a I I

+

I

b

2pV I

1 ms

A 6.0

iMD

PC~ ~\i

~ .

i \.

"

"

~CG~

.•.

~''.: INC .: Rr~ :.-.~01.6 ~

i

~VPM

J2i;/,

i /"

CG

:

i/

-FFH

i/;

0

<2 ~,A

:

O

2-10

!:" :!

O 10-20

IIA

o 20-30

~A

i\ :

.

. . . . :

.

! Cx:

,

/f

1 rnm

.

;,

~.-'~/i

:

: :i

"R i /.."

:

:

: "i-~.~

!

i

!" :

MMi \

.A

3o,,,< /:

ron discharged a n t i d r o m i c spikes with a short and fixed latency, following single shocks to the ipsilateral M L F (Fig. 6E) and the SR subdivision (Fig. 6G). The spikes followed suprathreshold-intensity double shocks with short intervals to the M L F (Fig. 6F) and the SR subdivision (Fig. 6H). The majority of neurons were m o r p h o logically verified to be in the superior vestibular nucleus, and only a few in the lateral vestibular nucleus. Figure 7 illustrates a result obtained by means of spike-triggered averaging of field potentials recorded in the SR subdivision. A n t i d r o m i c activation showed that the PC-related neuron projected to the SR subdivision

Fig. 8 Axonal projections of a PC-related inhibitory neuron to the upper brain stem (A 5.0, A 6.0, A 7.0 and A 8.0 Horsely-Clarke planes). The size of each circle indicates the threshold current for antidromic activation of the neuron as calibrated at right bottom. Numerals by the effective sites indicate antidromic latencies in ms. (CL central lateral nucleus, FR fasciculus retroflexus, MD medial dorsal nucleus, M M medial mamillary nucleus, PC posterior commissure, 3N oculomotor nerve, R red nucleus, VPM ventral posteromedial nucleus)

14

F

9

7.0 ~

(Fig. 7B). The field potential averaged with spikes of the neuron as the trigger (Fig. 7Ca) revealed a negative-positive spike followed by a positive wave (Fig. 7Cb). The succeeding positive wave had a monosynaptic latency of 0.9 ms, and was characterized by a steep rising phase and a relatively long recovery phase. Synaptic potentials in the SR subdivision were studied in 14 of the 26 inhibitory neurons. All the 14 inhibitory neurons induced the characteristic spike and positive wave complexes in the averaged field potentials recorded from the SR subdivision. Latencies of the positive wave measured from the triggering onset were ranging from 0.6 to 1.1 ms (mean 0.7 ms, n = 14). Electrophysiological studies of the axonal projection of single PC-activated inhibitory secondary vestibular neurons Axonal projections of single inhibitory PC neurons were investigated by antidromic threshold mapping in the mesodiencephalon on the ipsilateral side to the soma. Figure 8 shows the axonal projection of a representative inhibitory neuron. The effective sites were distributed in the ventral part of the central gray, INC, and its neighboring reticular formation, and rostrally in the mediocaudal part of the FFH. Low threshold foci were located in the dorsal and medial part of the INC. Latencies in the INC and ventral part of the central gray presented wide variations. In the ventral part of the INC, latencies were relatively prolonged. Eleven of the 26 inhibitory neurons were antidromi~cally activated from the INC and adjacent mesencephalic reticular formation. Figure 9 shows the effective foci in the 11 inhibitory neurons superimposed on three transverse planes through the levels of the INC, caudal, and rostral parts of the FFH. The inhibitory neurons were antidromically activated from limited areas including the central gray, INC, reticular formation surrounding them, and the caudalmost part of the FFH. Ten

A 8.0 ~

\

Fig. 9 Distribution of effectivesites in the mesodiencephalon for antidromic activation of 11 PC-related inhibitory neurons. Antidromic spikes were elicited in the hatched areas with stimuli less than 30 gA. The results of the 11 inhibitory neurons are superimposed in three transverse planes (A 6.0, A 7.0 and A 8.0 HorselyClarke planes). ( C M centromedian nucleus, F R fasciculus retroflexus, M D medial dorsal nucleus, M M medial mamillary nucleus, 3N oculomotor nerve, R red nucleus, VPM ventral posteromedial nucleus) other neurons showed no antidromic activation in any region rostral to the INC, when systematic tracking was made at 500-1000 t~m intervals with 50-250 gA stimulation in the region from A 7.0 to A 8.0 of the HorselyClarke plane, which were 8 mm lateral from the midline and 10 mm deep from the dorsal surface of the thalamus. In the remaining five neurons, antidromic activation was obtained in the SR subdivision, but not from more rostral areas.

Discussion Methodological consideration for identification In the present study, individual PC-related secondary vestibular neurons were identified by using natural and electrical stimulation of the labyrinth and antidromic activation following stimulation of the M L F and oculomotor nucleus. In cats, PC-related excitatory neurons always projected to the ipsilateral superior oblique and contralateral IR motoneurons through the contralateral M L F (Uchino et al. 1981; Uchino et al. 1982). When inhibitory neurons are connected with the antagonists of the above-noted motoneurons, axons travel in the ipsilateral M L F (Uchino et al. 1981; Uchino and Suzuki 1983). Therefore, PC-related secondary vestibular neurons antidromically activated from the contralateral M L F and the IR subdivision are presumably excitatory in nature, while the neurons antidromically activated from the ipsilateral M L F and the SR subdivision are

15 inhibitory. To identify the excitatory and inhibitory neurons, we further confirmed their monosynaptic connections with motoneurons by post-spike averaging of compound potentials of the IR and SR subdivisions triggered by spikes of PC-related neurons. The averaged field potentials, which were induced by PC-related secondary vestibular neurons antidromically activated from the contralateral M L F and the IR subdivision, had a biphasic spike followed by a negative wave in the IR subdivision. The initial spike probably represents the arrival of orthodromically conducted presynaptic impulses of the PC-related neuron (Nakao et al. 1982). The negative field potentials probably reflect excitatory postsynaptic currents and increased spike density of the IR motoneurons, because the negative wave had a steep falling phase and a relatively long recovery phase. The time course of the negative field potentials is compatible with that of unitary postsynaptic potentials in oculomotoneurons by single vestibular neurons (Uchino et al. 1981). A single premotor neuron increases the density of motoneuron spikes when mass discharges in the motor nerve are averaged by spike triggering (Nakao et al. 1982). When recordings are made within the nucleus of the motor nerve, the density of spikes increases as a negative deflection of the field potentials (Iwamoto et al. 1990). On the other hand, the averaged field potentials, which were induced by PC-related vestibular neurons antidromically activated from the ipsilateral M L F and the SR subdivision, showed a biphasic spike followed by a positive wave in the SR subdivision. The initial spike was presumably due to presynaptic impulses in the axon of the PC-related neuron (Nakao et al. 1982). The positive field potentials are known to reflect an extracellular counterpart of inhibitory postsynaptic currents and decreased density of spikes in the background noise (Iwamoto et al. 1990). Thus, PC-related excitatory and inhibitory neurons make direct connections with IR and SR motoneurons, and these neurons contribute, at least in part, to the vertical vestibuloocular reflexes.

Fig. 10A,B Schematic reconstruction of axonal branches of PC-activated secondary vestibular neurons in the mesodiencephalon. The crossing contralateral neuron is excitatory (A); the ipsilateral ascending neuron is inhibitory (B). (III oculomotor nucleus, D descending vestibular nucleus, L lateral vestibular nucleus, M medial vestibular nucleus, S superior vestibular nucleus, TH thalamus)

Projections to premotor areas in the mesodiencephalic junction Many single PC-related excitatory neurons were antidromically activated from the mesodiencephalic junction including the central gray, INC and FFH, and also from the thalamus in the present study (Fig. 10A). The effective sites for antidromic activation of each neuron were in the ventral part of the central gray, INC, and its neighboring reticular formation. Low threshold foci were present between the INC and central gray, indicating that the stem axon may travel into this region. The INC and ventral part of the central gray showed wide variations in thresholds and latencies, which suggests that the neuron gives off collaterals that terminate in these regions. The effective sites for antidromic activation within and around F F H were located in the ventral part of the central gray, F F H and zona incerta. Antidromic spike latencies and thresholds varied widely among the effective sites in the medial part of the FFH. The latencies were shorter in the medial part of the F F H and central gray than in the lateral part of the FFH. These observations suggest that the PC-related excitatory neuron issues numerous axon branches, which terminate at least in the medial part of the FFtt, traveling to the zona incerta in the mediolateral direction. The INC is presumed to be a vertical gaze center, playing an integrating role in changing vertical eye velocity signals into position signals (Fukushima 1991). The F F H of cats plays a similar role to that of the riMLF of monkeys (Biittner-Ennever and Biittner 1978) and is known to be an important premotor area for vertical saccadic eye movements. Neurons in the cat F F H exhibit various patterns of activities related to vertical saccades (King and Fuchs 1979; Nakao and Shiraishi 1985). In the FFH, medium-lead burst neurons are mainly located in the medial part, long-lead burst neurons in the caudal part. On the other hand, tonic neurons are mostly in the INC and central gray caudal to the FFt-I, and burst-

I I III ,

i

itM

Illl

16

tonic neurons are within and near the INC (Nakao et al. 1986). These locations almost correspond to the projection areas of PC-related excitatory neurons in the present study. There are anatomical similarities in the axons of PC-related secondary vestibular neurons projecting to the INC between cats (Graf et al. 1983, 1986; Iwamoto et al. 1990) and monkeys (McCrea et al. 1987). Neurons in the INC receive disynaptic inputs from vestibular vertical canal nerves (King et al. 1980; Fukushima et al. 1991). Tonic and burst-tonic neurons in and around the INC are monosynaptically activated by contralateral vestibular nucleus stimulation (Nakao et al. 1987; Fukushima et al. 1991). In the FFH, medium-lead burst neurons are activated from the ipsilateral or contralateral vestibular nucleus; and long-lead burst neurons are activated by the vestibular nucleus with monosynaptic or disynaptic latencies (Nakao et al. 1987). These findings suggest that the PC-related neurons make direct or indirect excitatory connections with various mesodiencephalic neurons related to vertical eye movements. The projections of excitatory neurons to the premotor areas shown in the present study were classified into two types depending on whether the neuron was with or without axon collaterals to the FFH. In order to clarify whether the two types of neurons are functionally different, it is important to determine the connections between these two types of neurons and the functionally identified neurons in the mesodiencephalic junctions. Conversely, the PC-related inhibitory neurons projected mainly to the INC and caudal part of the F F H (Fig. 10B), suggesting that these neurons project to premotor neurons in the caudal mesodiencephalic junction.

suprasylvian sulcus in cats (Liedgren et al. 1976; Blum et al. 1979; Mergner et al. 1981). The main locations of these neurons are reported to be in the ventral posterolateral nucleus, posterior margin of the ventrobasal complex, and magnocellular nucleus of the medial geniculate body. Similarly, as we observed in the present study, many axon branches of PC-related excitatory secondary vestibular neurons were in the ventrobasal complexes of the thalami. The axon branches probably contribute to the primary vestibulo-cortical pathways. We did not find PC-related neurons that clearly projected to the magnocellular nucleus of the medial geniculate body. The centromedian nucleus is suggested to be an important element in the loop system that links the medial thalamus and neostriatum (McGuinness and Krauthamer 1980), and the ventral lateral geniculate nucleus may be an important nucleus for integrating vestibular and visual information in the performance of head and eye movements (Magnin and Kennedy 1979). Some axon branches of PC-related excitatory secondary vestibular neurons studied here projected also to the centromedian, central lateral, and ventral lateral geniculate nuclei. These thalamic nuclei probably receive excitatory input from PC-related secondary vestibular neurons. Since PC-related inhibitory secondary vestibular neurons gave off few axons to the thalamus, these inhibitory neurons do not appear to contribute to the vestibulo-thalamic pathways. We conclude that PC-related excitatory secondary vestibular neurons participate in the genesis of vertical eye movements and in the perception of the vestibular sensation, while PC-related inhibitory secondary vestibular neurons seem to take part only in the genesis of vertical eye movements.

Projections to the thalamus Many PC-related excitatory neurons were antidromically activated from the thalamic nuclei, particularly the ventral posteromedial and ventral posterolateral nuclei. They have vastly more projections to the ventral part of lateral posterior, central lateral, centromedian, ventral lateral, ventral medial, and ventral lateral geniculate nuclei (Fig. 10A). The antidromic latencies for each excitatory neuron were longer in the lateral sites, indicating that excitatory neurons may give off axon collaterals in the thalamic nuclei, which course in the mediolateral direction. Cortical areas which receive input from the vestibular nerve are located in the postcruciate dimple and anterior suprasylvian sulcus in cats (Landgren et al. 1967; Sans et al. 1970; Odkvist et al. 1975; Liedgren and Schwartz 1976). These areas receive convergent inputs from proprioceptive, somatosensory, and vestibular afferents via thalamic neurons, and were thereby postulated to contribute to conscious spatial orientation (Fredrickson et al. 1966; Sans et al. 1970; Schwartz and Fredrickson 1971). Thalamo-cortical neurons were studied by retrograde transport of horseradish peroxidase injected into the postcruciate dimple and anterior

Acknowledgements The authors want to thank Dr. Kawai for helpful criticism of the manuscript. This study was partly supported by a Grant-in-Aid for neuroscience research from a 1991 Narishige Research Fund.

References Blum PS, Day MJ, Carpenter MB, Gilman S (1979) Thalamic components of the ascending vestibular system. Exp Neurol 64:587-603 B/ittner U, Bfittner-Ennever JA, Henn V (1977) Vertical eye movement related unit activity in the rostral mesencephalic reticular formation of the alert monkey. Brain Res 130:239-252 B/ittner-Ennever JA, Biittner U (1978) A cell group associated with vertical eye movements in the rostral mesencephalic reticular formation of the monkey. Brain Res 151:31-47 Fredrickson JM, Figge U, Scheid P, Kornhuber HH (1966) Vestibular nerve projection to the cerebral cortex of the rhesus monkey. Exp Brain Res 2:318-327 Fukushima K (1991) The interstitial nucleus of Cajal in the midbrain reticular formation and vertical eye movement. Neurosci Res 10:159-187 Fukushima K, Suzuki Y, Fukushima J, Kase M (1991) Latencies of response of eye movement-related neurons in the region of the interstitial nucleus of Cajal to electrical stimulation of the vestibular nerve in alert cats. Exp Brain Res 87:254-258

17 Graf W, Ezure K (1986) Morphology of vertical canal related second order vestibular neurons in the cat. Exp Brain Res 63: 35.48 Graf W, McCrea RA, Baker R (1983) Morphology of posterior canal related secondary vestibular neurons in rabbit and cat. Exp Brain Res 52:125 138 Graybiel AM (1977) Organization of oculomotor pathways in the cat and rhesus monkey. In: Baker R, Berthoz A (eds) Control of gaze by brain stem neurons. Developments in neuroscience, vol 1. Elsevier, New York, pp 79-88 Isu N, Sakuma A, Kitahara M, Uchino Y, Takeyama I (1991) Vestibulo-thalamic neurons give off descending axons to the spinal cord. Acta Otolaryngol [Suppl 481]:216-220 Iwamoto Y, Kitama T, Yoshida K (1990) Vertical eye movementrelated secondary vestibular neurons ascending in medial longitudinal fasciculus in cat II. Direct connections with extraocular motoneurons. J Neurophysiol 63:918-935 Jankowska E, Roberts WJ (1972) An electrophysiological demonstration of the axonal projections of single spinal interneurons in the cat. J Physiol (Lond) 222:597-622 King WM, Fuchs A F (1979) Reticular control of vertical saccadic eye movements by mesencephalic burst neurons. J Neurophysiol 42:861-876 King WM, Precht W, Dieringer N (1980) Synaptic organization of frontal eye field and vestibular afferents to interstitial nucleus of Cajal in the cat. J Neurophysiol 43:912-928 K6mpf D, Pasik T, Pasik P, Bender MB (1979) Downward gaze in monkeys. Stimulation and lesion studies. Brain 102:527-558 Kotchabhakdi N, Rinvik E, Walberg F, Yingchareon K (1980) The vestibulothalamic projections in the cat studied by retrograde axonal transport of horseradish peroxidase. Exp Brain Res 40:405-418 Lang W, Biittner-Ennever JA, Biittner U (1979) Vestibular projections to the monkey thalamus: an autoradiographic study. Brain Res 177:3-17 Landgren S, Silfvenius H, Wolsk D (1967) Vestibular, cochlear and trigeminal projections to the cortex in the anterior suprasylvian sulcus of the cat. J Physiol (Lond) 191:561-573 Liedgren SRC, Schwartz D W F (1976) Vestibular evoked potentials in thalamus and basal ganglia of the squirrel monkey (saimiri sciureus). Acta Otolaryngol 81:73-82 Liedgren SRC, Kristensson K, Larsby B, ()dkvist LM (1976) Projection of thalamic neurons to cat primary vestibular cortical fields studied by means of retrograde axonal transport of horseradish peroxidase. Exp Brain Res 24:237-243 Magnin M, Kennedy H (1979) Anatomical evidence of third ascending vestibular pathway involving the ventral lateral geniculate nucleus and intralaminar nuclei of the cat. Brain Res 171 : 523-529 Matsuo S, Hosogai M, Nakao S (1992) Axonal projection in mesodiencephalon of cat posterior canal-activated inhibitory secondary vestibular neurons. Neurosci Res [Suppl 17]:$210 McCrea RA, Strassman A, Highstein SM (1987) Anatomical and physiological characteristics of vestibular neurons mediating the vertical vestibulo-ocular reflexes of the squirrel monkey. J Comp Neurol 264:571-594

McGuinness CM, Krauthamer GM (1980) The afferent projections to the centrum medianum of the cat as demonstrated by retrograde transport of horseradish peroxidase. Brain Res 184: 255-269 Mergner T, Deeke L, Wagner HJ (1981) Vestibulo-thalamic projection to the anterior suprasylvian cortex of the cat. Exp Brain Res 44:455.458 Nagata S (1986) The vestibulothalamic connections in the rat: a morphological analysis using wheat germ agglutininhorseradish peroxidase. Brain Res 376:57-70 Nakao S, Shiraishi Y (1985) Direct excitatory and inhibitory synaptic inputs from the medial mesodiencephalic junction to motoneurons innervating extraocular oblique muscles in the cat. Exp Brain Res 61:62-72 Nakao S, Sasaki S, Schor RH, Shimazu H (1982) Functional organization of premotor neurons in the cat medial vestibular nucleus related to slow and fast phases of nystagmus. Exp Brain Res 45:371-385 Nakao S, Shiraishi Y, Oikawa T (1986) Vertical eye movement-related neurons in the medial mesodiencephalic junction: their firing pattern, location and projection to oculomotor and trochlear nuclei. Neurosci Res [Suppl 3] :$66 Nakao S, Shiraishi Y, Oikawa T (1987) Vertical eye movement-related neurons in the cat mesodiencephalic junction: synaptic input from or output to the vestibular nucleus. Neurosci Res [Suppl 5] :$73 Odkvist LM, Liedgren SRC, Larsby B, Jerlvall L (1975) Vestibular and somatosensory inflow to the vestibular projection area in the post cruciate dimple region of the cat cerebral cortex. Exp Brain Res 22:185-196 Ohgaki T, Markham CH, Schneider JS, Curthoys IS (1989) Anatomical evidence of the projection of pontine omnipause neurons to midbrain regions controlling vertical eye movements. J Comp Neurol 289:610-625 Sans A, Raymond J, Marty R (1970) Reponses thalamiques et corticales a la stimulation electrique du nerf vestibulaire chez le chat. Exp Brain Res 10:265-275 Schwartz DWF, Fredrickson JM (1971) Rhesus monkey vestibular cortex: a bimodal primary projection field. Science 172:280-281 Shimazu H, Precht W (1965) Tonic and kinetic responses of cat's vestibular neurons to horizontal angular acceleration. J Neurophysiol 28:991-1013 Thomas RC, Wilson VJ (1965) Precise localization of Renshaw cells with a new marking technique. Nature 206:211~13 Uchino Y, Suzuki S (1983) Axon collaterals to the extraocular motoneuron pools of inhibitory vestibulo-ocular neurons activated from the anterior, posterior and horizontal semicircular canals in the cat. Neurosci Lett 37:129-135 Uchino Y, Hirai N, Suzuki S, Watanabe S (1981) Properties of secondary vestibular neurons fired by stimulation of ampullary nerve of the vertical, anterior or posterior, semicircular canals in the cat. Brain Res 223:273 286 Uchino Y, Hirai N, Suzuki S (1982) Branching pattern and properties of vertical- and horizontal-related excitatory vestibuloocular neurons in the cat. J Neurophysiol 48:891-903