Brain Research, 148 (1978) 31--42 © Elsevier/North-HollandBiomedicalPress
31
VESTIBULAR NUCLEUS NEURONS RELAYING EXCITATION FROM THE ANTERIOR CANAL TO THE OCULOMOTOR NUCLEUS
M. YAMAMOTO, I. SHIMOYAMA and S. M. HIGHSTEIN University of Tokyo, Faculty of Medicine, Department of Physiology, Hongo, Bunkyo-Ku, Tokyo 176 (Japan) and (S.M.H.) Department of Neuroscience, Albert Einstein College of Medicine, Bronx, N. Y. 10461 (U.S.A.)
(Accepted September 22nd, 1977)
SUMMARY A morphophysiological approach was undertaken to determine which vestibular nucleus neurons relay excitation from the anterior canal to the IIIrd nucleus. In anesthetized rabbits HRP was iontophoresed into the IIIrd nucleus and cells filled with HRP reaction product (positive cells) searched for within the vestibular nuclear complex. By lesioning the MLF or brachium conjunctivum immediately after iontophoresis it was demonstrated that positive cells in the dorsum of the superior vestibular nucleus are backfilled via their axons which ascend in the brachium conjunctivum. By COntrast positive cells in the center of the superior nucleus are backfilled via their axons in the MLF. In electrophysiological experiments in the presence of a severed MLF the anterior canal was selectively stimulated' for orthodromic~ and the 3rd nucleus stimulated for antidromic, activation of vestibular nucleus neurons. Recording extracellularly with glass microelectrodes filled with fast green FCF the only ceils both orthoand 'antidromically activated were localized to the dorsum of the superior vestibular nucleus. It is concluded that cells dorsally located in the superior nucleus relay the disynaptic excitatory vestibulo-ocular reflex from the anterior canal to the contralateral 3rd nucleus via their axons which ascend in the brachium conjunctivum.
INTRODUCTION The purpose of this paper is t o determine which cells in the vestibular nuclei relay anterior canal excitation to the 3rd nucleus. There is physiological1~and morphologicalS, 7 evidence indicating that labyrinthine signals to oculomotor neurons are conveyed not only through the medial longitudinal fasciculus (MLF) but also via the brachium conjunctivum (BC). The presumed cells of the origin" of the BC pathway
32 to several subgroups of neurons in the third nucleus were located with a microstimulation technique to the region of the Y group of the vestibular nuclear complex and to the dentate nucleus of the cerebeUumL Both the Y group and the dentate nucleus receive primary vestibular afferent inputsa, 4, and it has been asserted that both project directly to the oculomotor nuclear complexS,7, 9. Anterior canal stimulation evokes reflex contraction in the ipsilateral superior rectus and the contralateral inferior oblique extraocular muscles in the rabbit 16. The BC pathway relays these anterior canal signals which remain after aspiration of the dentate nucleus 16. As it was suggested 9 that the dentate nucleus and the Y group contain the ceils of origin of the BC pathway it seemed logical to deduce that the anterior canal excitation is mediated by the Y group 16. However, Gacek 6 has suggested that direct afferent input to the Y group arises from the saccule and not from the utricle or semicircular canals. In contrast to Gacek's morphological results 6, Hwang and Poon ~a recorded from the Y group during stimulation of the saccule and found only multisynaptic activation of Y group neurons. In order to unravel this puzzle of which cells relay the anterior canal excitation to the 3rd nucleus a morphophysiological approach was undertaken. Our results show that there is a previously unnoticed population of ceils in the dorsal portions of the superior vestibular nucleus which project through the BC to the IIIrd nucleus. It is these superior nucleus cells which are now presumed to be the anterior canal relay neurons. MATERIALS AND METHODS
Morphological experiments Nine albino rabbits were anesthetized with intravenous urethane (400 mg/kg) and a-chloralose (40 mg/kg) and were mounted in a stereotaxic apparatus. Craniotomy was performed and the occipital lobe on the right side was aspirated to expose the surface of the superior colliculus. The whole 3rd nerve was stimulated stereotaxically 12. A glass microelectrode (2 M NaC1, resistance 2-4 Mf~) was inserted through the superior colliculus to identify the region of the 3rd cranial nucleus by observing the antidromic field potentials evoked by 3rd nerve stimulation. Once the limits of the nucleus were defined the recording electrode was replaced with a bevelled glass microelectrode filled with concentrated horseradish peroxidase (HRP) solution in Tris buffer pH 8.6. We attempted to place the H R P electrode 1 mm rostral to the caudal edge of the 3rd nucleus. H R P was iontophoresed using 1-3/~A 250 msec positive pulses every 500 msec for a total on time of 30-60 minS, 22. In 5 animals, immediately after the iontophoretic procedure the cerebellum was exposed and the vermal portions aspirated to expose the floor of the 4th ventricle. A transverse incision in the brain stem at a level several millimeters anterior to the superior vestibular nucleus was made. The intent of these lesions was to sever the median longitudinal fasciculus (MLF) and the pontine tegmentum bilaterally. As the pathways relaying disynaptic vestibuloocular signals have been demonstrated to be the M L F 7,1°,12,15, BC 5,1°,1g,16 and ascending tract of Deitersll the intent of our lesions was to interrupt
33 these pathways. Even though our lesions often impinged upon reticular structures we felt that this did not interfere with the purpose of this study, as reticular nuclei have not been implicated in the disynaptic vestibuloocular reflexes. In one rabbit the BC was interrupted by a transverse incision just posterior to the base of the inferior colliculus. In the remaining 3 animals the brain stem was left intact. Twenty to 24 h after the injection the animals were perfused with 10 ~ formalin in 0.1 M phosphate buffer at pH 7.4. The brains were removed from the skull and stored in cold phosphate buffer containing 30 ~ sucrose. On the following day serial frozen sections, 50/~m thick were cut, incubated for 30-60 min in hydrogen peroxide and 3,3'-diaminobenzidine tetrahydrochloride. The sections were then mounted on glass slides counter stained with 0.5~ cresyl violet, and drawn at 10 x magnification. Subsequently sections were examined with a light microscope and the locations of cells labelled with HRP reaction product were marked on the drawings.
Electrophysioiogical experiments Fifteen rabbits were used, anesthetized with chloralose-urethane. The trachea was cannulated, the rabbits were placed in a stereotaxic apparatus and artificially respired. Through a dorsal approach the bony labyrinth was visualized and the area overlying the ampulla of the anterior canal identified. A small hole was made in the bone overlying the ampulla and isolated stimulation of the anterior canal with fine bipolar wires was attempted 14. EMG electrodes were placed in the ipsilateral superior rectus and medial rectus extraocular muscles. The threshold for EMG activation of the superior rectus muscle via anterior canal stimulation was determined. Because the horizontal canal nerve is the most likely vestibular structure to be excited by current spread during anterior canal stimulation and it is known that the horizontal canal relays excitation to the medial rectus muscle, we determined how many multiples of threshold for superior rectus activation were necessary before medial rectus activation was observed. During the subsequent electrophysiological experiments, care was taken to insure that the stimulus intensity to the anterior canal was always below that which evoked excitation in the medial rectus muscle. Craniotomy was performed, the occipital cortex aspirated and a concentric bipolar stimulating electrode inserted through the superior colliculus to the region of the oculomotor nuclear complex. Final position of this stimulating electrode was adjusted by recording the field potentials following anterior canal stimulation. The cerebellum was exposed by craniotomy, its vermal portions aspirated and a broad deep cut of the pons rostral to the superior nucleus was made to interrupt the medial longitudinal fasciculus and the pontine tegmentum bilaterally. Single units in the area of the vestibular nuclei were isolated with a recording microelectrode filled with fast green FCF 2a during stimulation of the anterior canal for orthodromic and the oculomotor nucleus for antidromic activation. Sites of positive cells anti- and orthodromically activated were marked with fast green and were subsequently identified and indicated on histological reconstructions of brain stem sections.
34
A
3,d..c
H
~ - -
~ o
¢ 3,
D _~.~ ~ ~/~-.'.~> '~~
I
~/~
Fig. 1. H R P injection into the Iltrd nucleus with an intact brain stem. A - C illustrate the injection site. The black area indicates the region most heavily impregnated with reaction product, the shaded area is a zone of lighter, but still appreciable impregnation and the dotted lines indicate the limit of visible H R P diffusion. The most heavily impregnated area was limited to the right oculomotor nucleus. D - K illustrate the distribution of H R P positive cells in the brain stem. BC, brachium conjunctivum; DV, descending vestibular nucleus; DN, lateral cerebellar nucleus; LV, lateral vestibular nucleus; MV, medial vestibular nucleus; RB, restiform body; SV, superior vestibular nucleus; Y, vestibular Y group; III, oculomotor nucleus; VI, abducens nucleus.
35
Fig. 2. Light micrograph of HRP positive cells within the dorsal portions of the superior nucleus. Fine granules can be seen. The bar indicates 20 #m. RESULTS
(A) Morphology (1) Superior nucleus cells projecting through the brachium conjunctivum to the third nucleus (a) Control animals. The 3 intact animals had similar results; Fig. 1 illustrates a typical case. Nerve cells in and around the vestibular nuclei which contain horseradish peroxidase granular reaction product were regarded as positive cells (Fig. 2). Diffusely stained cells were discarded as they might have been artifactually stained due to interruption of their axons by the pipettes placed in the 3rd nucleus. In the case illustrated in Fig. 1, the dense HRP reaction product was mainly localized to the right oculomotor nuclear complex. The size of the injection site was defined as the limit of the spread of the homogeneous horseradish reaction product within the 3rd nucleus. Iontophoretic application of HRP in general yielded a restricted spread of label at the injection site. The site consisted of a central region which was densely stained with reaction product surrounded by a much more lightly stained zone. The limits of the site can be visualized in Fig. 1A, B and C. The dense black area within the oculomotor nuclear complex is the area of dense reaction product. The cross hatched area indicates the more lightly labeled zone and the dotted line indicates the limits of the visible reaction product. It is likely that only cells which project to the densely labeled portions of the injection site retrogradely transport horseradish peroxidase 22. In sections D through H, there are labeled cells in the superior nuclei bilaterally. In both the Y group (1I and J) and the superior nucleus positive cells are either spindle shaped with flattened nuclei or multipolar with round nuclei. The cells appear to lie in rows between the passing fibers. The size of these ceils is approximately 20-50 #m which is comparable to the diameters of cells within the center of the nucleus. Typical cells are illustrated in Fig. 2. In Fig. 1F the distribution of labeled cells within the bilateral superior nucleus is of interest. Contralateral to the injection site the labeled cells are concentrated within the dorsal parts of the superior nucleus, the central
36
A
B
Fig. 3. HRP injection into the IIIrd nucleus with a severed MLF and pontine tegmentum. Labeling as in Fig. 1. A indicates the injection site. B indicates the extent of the lesion of the brain stem (black area). In this case the injection was unilateral and only the contralateral superior nucleus contains positive cells.
portions of the nucleus being free of positive ceils. On the side ipsilateral to the injection site the dorsal portions of the nucleus are free of labeled cells, while the labeled ceils are concentrated in the central regions of the nucleus. The contralateral medial vestibular nucleus ( 1 H - K ) contains labeled ceils in its rostral portions but there are also labeled cells ipsilaterally (1 H). In addition the contralateral descending vestibular nucleus contains positive cells (1J and K). The Y group overlying the restiform body is clearly filled with many positive cells (1I and J). Additionally there are positive cells lying within the contralateral abducens nucleus (1I). These are presumably abducens internuclear neurons projecting to the contralateral medial rectus subgroup as has been described in the cat2,7,11. Cell bridges between the Y group and the dentate nucleus also contain positive cells. There were no positive cells in either the dentate nucleus of the cerebellum or in dorsal Deiter's nucleus. There were two additional
37
A
Fig. 4. HRP injection into the IIIrd nucleus with a severed BC. A shows the injection sites. B shows the extent of the BC lesion. Note that in this case the injection was bilateral and the left BC was cot. animals in the control group. In one animal the injection site covered the whole third nucleus bilaterally and the lightly stained area extended to the MLF. Positive cells in this case were found mainly in the medial and the superior vestibular nuclei. In the third case the injection site was small and localized within the caudal half of the bilateral third nuclei. Though the general features of the distribution of positive cells were similar to that in the first two cases the number of labeled cells were less abundant in the medial nucleus and descending nucleus. (b) Lesion experiments. In 5 rabbits horseradish peroxidase was iontophoresed into the 3rd nucleus unilaterally and the M L F was interrupted bilaterally. The extent of the lesion is exemplified in Fig. 3B. In all 5 of these cases positive cells
38
F
---
G
Fig. 5. Electrophysiologicai analysis of the distribution of neurons orthodromically activated by anterior canal stimulation and/or antidromically activated by llIrd nucleus stimulation. A: orthodromic activation of a neuron by anterior canal stimulation and B: antidromic activation of the same neuron by stimulation at the site marked in black in D. C: reproduction of Fig. IF. D and E: cross sections of the brain stem to indicate the site of stimulation within the IIIrd nucleus in D and the extent of the brain stem lesion in E. F - H : brain stem sections to indicate the location of cells orthodromically activated by anterior canal and antidromically activated by IIIrd nucleus stimulation (~k). Orthodromic only cells indicated by O; antidromic only cells indicated by 0 .
were seen in the c o n t r a l a t e r a l b u t n o t in the ipsilateral Y g r o u p ( 3 D - F ) . In a d d i t i o n some cells were f o u n d within the dorsal p o r t i o n s o f the c o n t r a l a t e r a l superior vestibular nucleus ventral to the BC a n d medial to the restiform b o d y (3C a n d D). I n a n o t h e r r a b b i t the BC was i n t e r r u p t e d unilaterally after the i o n t o p h o resis o f p e r o x i d a s e into the o c u l o m o t o r nucleus. In this case the i o n t o p h o r e t i c inject i o n e n c o m p a s s e d the c a u d a l two-thirds o f b o t h third nuclei (Fig. 4A). O n the side o f the intact BC cells in the Y g r o u p were labeled j u s t as described in the previous five p r e p a r a t i o n s ( 4 E - G ) . O n the lesion side however, there were no positive cells
39 within the Y group (Fig. 4). The distribution of labeled cells in the bilateral superior nuclei, in this case with a lesioned BC, complements the cases with M L F lesions. On the intact side positive cells are found throughout the superior nucleus (4C-E). That is, in its central portions as well as in the dorsal portions underlying the BC. However, on the lesion side only the central and caudal portions of the superior nucleus are stained (4C-E). The dorsal portions of the nucleus underlying the BC are not filled with positive cells. This can be seen quite clearly by comparing Fig. 4C and 3C. This implies that those positive cells underlying the BC are labeled via the BC, while those cells in the center of the nucleus are labeled via the M L F pathway. In another series of rabbits, both the BC and M L F were extensively interrupted on both sides. In this case, no positive cells were found anywhere within the vestibular nuclei. This control experiment eliminates the possibility that the above described results were an artifact of either endogenous peroxidases or of endogenous peroxidase uptake. ( B) Electrophysiological identification o f cells activated by anterior canal and oculomotor nuclear stimulation
After measuring the midline and the depth of the floor of the fourth ventricle with a microelectrode, systematic tracking was made throughout the area of the vestibular nuclear complex. Maps were drawn of the tracks indicating cells which were activated antidromically by third nucleus or orthodromicaUy by anterior canal stimulation. Fig. 5A and B illustrate the extracellular records taken from a unit which was both orthodromically (5A) and antidromically (5B) activated. Orthodromic activation times were typically 1-1.5 msec. There was some jitter in the latency to activation as expected with orthodromic activation. It has been reported 19 that superior nucleus neurons can be activated in as little as 0.8 msec following supramaximal eighth nerve stimulation. However, in the present experiment stimulation intensities employed were rather weak in order to insure isolated activation of the anterior canal. Thus the latencies to activation (1-1.5 msec) probably lie within the monosynaptic range. Antidromic activation was all-or-none with fixed latency. The typical latency for antidromic activation was TABLE I The distribution of electrophysiologically identified cells in the rabbit's brain stem
Abbreviations: AC, anterior canal; A and O, antidromically and orthodromically activated cells; A, Cells activated only antidromically; O, Cells activated only orthodromically; LV, Deiter's nucleus; DN, dentate nucleus of the cerebellum; SV, superior vestibular nucleus; VLD, ventrolateral vestibular nucleus; NVC, nucleus cerebello-vestibulosis; BC, brachium conjunctivum; SV-Y, area between the Y group and SV; SV-BC, area between the BC and the SV; Y, Y group. The numbers indicate the number of cells identified as lying within the structures indicated in each column. AC
O-}-A A O
LV
23
DN
SV
VLD
NVC
BC
SV- Y
SV-BC
17 14 52
1 2
2
4
5 19 12
2
2
Y
7 1
40 0.5-0.6 msec. When double shock stimulation was given to a unit which was antidromically activated, the latency to the activation following the second stimulus was often significantly longer than the first (not illustrated). This is presumptive evidence for antidromic activation. Cells were counted as being activated by the anterior canal if they could be activated with low stimulus intensity as mentioned in the Methods. The electrophysiological results herein reported are based upon 16l units. Many other units were discarded as they could not be accurately located on the histological reconstructions. Fig. 5 illustrates the locations of units from 4 experiments. It can be seen (5F-H) that units which are ortho- and antidromically activated (~r) tend to lie in the dorsal parts of the superior nucleus, while units which are only orthodromically activated (O) lie either among the incoming fibers of the eighth nerve or in the more ventral portions of the superior nucleus. Records of the two units anti- and orthodromically activated within the BC (5F) were presumably taken from the dendrites of superior nucleus cells. Several units only antidromicaUy activated lie within the lateral nucleus of the cerebellum or adjacent to it in the interpositus. Fig. 1F is reproduced in 5C for comparison with the electrophysiological results (see Discussion). The distribution of 161 cells which were ortho- and/or antidromically activated is illustrated in Table I. DISCUSSION The present results support the previous electrophysiological evidence of a direct vestibular oculomotor projection through the BC. The rostral dorsal portions of the superior nucleus were demonstrated electrophysiologically to contain cells which carry signals from the anterior canal relaying them through the BC to the contralateral third nucleus. Morphological experiments support this view. It is striking to note the distribution of labeled cells in Fig. 1F (and 5C) within the bilateral superior nuclei. In this case the injection site is limited to the right oculomotor nucleus. Results herein presented in Figs. 3-5 suggest that the dorsally located cells in the left (contralateral to the injection) superior nucleus were labeled with HRP retrogradely transported via their axons in the BC while the centrally located cells in the right (ipsilateral) superior nucleus were labeled via their M L F projection. This view gains credence because the right third nucleus contains the motoneurons innervating the right inferior oblique and the left superior rectus 9, both of which are activated by left anterior canal stimulation 16. In this context it is also noted that this identical distribution of labeled neurons has been demonstrated in the cat by Graybiel and Hartweig 7 in their Fig. 1. That is, in Graybiel's paper if the injection of peroxidase is limited to one oculomotor nuclear complex (as in Fig. 1), then the identical distribution of cells in the bilateral superior nuclei as described for the rabbit is observed. However, when the injection encompassed both third nuclei (Graybiel and Hartweig, Fig. 2) the distinctive distribution of cells in the superior nuclei is lost. It can now be suggested that the oculomotor output of the superior nucleus is divided into at least two portions. The portion containing cells which lie directly underneath the BC is apparently relaying excitatory signals through the brachium
41 to the contralateral oculomotor nuclear complex. The superior nucleus cells relaying inhibition up the ipsilateral M L F 1°,12,15 lie within the central portions of the nucleus. It was previously thought that the cells in the periphery of the superior nucleus did not receive any input from the labyrinth 4. Our data suggests that the dorsal portions of superior nucleus receive primary input from the anterior canal. Gacek 6 also supports the view that the entire superior nucleus receives afferents from the labyrinth. Further our results (Fig. 2) indicate that neurons dorsally located within the nucleus can be as large as those in the center. Gacek (personal communication) also finds large ceils dorsally located. Secondly, flocculus Purkinje ceils were shown to end predominantly in the center of the superior nucleus 1. Since the flocculus inhibits the excitatory pathway from the anterior canal xT, we infer that it must also terminate on the dorsum of the superior nucleus. Alley (personal communication) substantiates this view utilizing labeled amino acids to map flocculus projections. The area of the Y group contains many H R P positive ceils after third nucleus injection as do the cell bridges between the Y group and dentate nucleus of the cerebellum. This area of cell bridges appears identical to the area which Lorente de N6 ~1 originally called the nucleus cerebellovestibulosis in the guinea pig. Gacek (personal communication) also confirms that the primary afferents from the saccule project to the area of these cell bridges. Y group cells were surveyed with a microelectrode but were not activated by isolated stimulation of the anterior canal. (Ito et al., unpublished; Yamamoto et al., this paper, Table I). Evidence to date6,13, 21 supports the idea that the Y-group is the relay nucleus for the sacculoocular reflex. Ceils projecting to the third nucleus were not found in the dentate nucleus of the cerebellum. Graybiel et al. 7 also failed to define labeled cells within the dentate except for a very few. Whether the failure of labeling in cats and rabbits indicates the absence of dentate ceils projecting to the third nucleus is still open to question. Experiments on this point are currently in progress. Lastly abducens internuclear neurons demonstrated in the cat 2,11,7 are found in the present study thus confirming this important pathway in the rabbit, and suggesting some common features in the organization of horizontal as well as vertical gaze. ACKNOWLEDGEMENTS The authors would like to thank Prof. M. Ito for his critical reading of the manuscript and for his support during this study. This work was supported by Grants No: N I H 5 K04 EY 00003-02, N I H 5 R01 EY 01670-02 and NSF OIP 74-13621.
REFERENCES 1 Angaut, P. and Brodal, A., The projection of the 'vestibulo-cerebellum' onto the vestibular nuclei in the cat, Arch. itaL BioL, 105 (1965) 441-479. 2 Baker, R. and Highstein, S. M., Physiological identification of interneurons and motoneurons in the abducens nucleus, Brain Research, 91 (1975) 292-298.
42 3 Brodal, A. and Hoivik, B., Site and mode of termination of primary vestibulocerebellar fibers in the cat, Arch. ital. Biol., 102 (1964) 1-21. 4 Brodal, A. and Pompeiano, O., The vestibular nuclei in the cat, J. Anat. (Lond.), 91 (1957) 438454. 5 Carpenter, M. B. and Strominger, N. C., Cerebello-oculomotor fibers in the rhesus monkey, J. comp. Neurol., 123 (1964) 211 230. 6 Gacek, R., The course and central termination of first order neurons supplying vestibular end organs in the cat, Acta Otolaryng. (Stockh.), Suppl., 254 (1969) 1-66. 7 Graybiel, A. M. and Hartwieg, E. A., Some afferent connections of the oculomotor complex in the cat: an experimental study with tracer techniques, Brain Research, 81 (1974) 543-551. 8 Graybiel, A. M. and Devor, M., A microelectrophoretic delivery technique for use with horseradish peroxidase, Brain Research, 68 (1974) 167-173. 9 Highstein, S. M., The organization of the vestibulo-oculomotor and trochlear reflex pathways in the rabbit, Exp. Brain Res., 17 (1973) 285-300. 10 Highstein, S. M., Organization of the inhibitory and excitatory vestibulo-ocular reflex pathways to the third and fourth nuclei in rabbit, Brain Research, 32 (1971) 218-224. 11 Highstein, S. M. and Baker, R., Termination of internuclear neurons of the abducens nuclei on medial rectus motoneurons, Neuroscience Abs., 2 (1976) 398. 12 Highstein, S. M., lto, M. and Tsuchiya, T., Synaptic linkage in the vestibulo-ocular reflex pathway of rabbit, Exp. Brain Res., 13 (1971) 306-326. 13 Hwang, J. C. and Poon, W. F., An electrophysiological study of the sacculo-ocular pathways in cats, Jap. J. Physiol., 25 (1975) 241-251. 14 Ito, M., Nisimaru, N. and Yamamoto, M., The neural pathways mediating reflex contraction of extraocular muscles during semicircular canal stimulation in rabbits, Brain Research, 55 (1973) 183-188. 15 lto, M., Nisimaru, N. and Yamamoto, M., The neural pathways relaying reflex inhibition from semicircular canals to extraocular muscles of rabbits, Brain Research, 55 (1973) 189-193. 16 Ito, M., Nisimaru, N. and Yamamoto, M., Pathways for the vestibulo-ocular reflex excitation arising from semicircular canals of rabbits, Exp. Brain Res., 24 (1976) 257-271. 17 Ito, M., Nisimaru, N. and Yamamoto, M., Specific patterns of neuronal connections involved in the control of rabbits' vestibulo-ocular reflexes by the cerebellar flocculus. J. Physiol. (Lond.), 265 (1977) 833-854. 18 lto, M., Nisimaru, N. and Yamamoto, M., Unpublished. 19 Kawai, N., Ito, M. and Nozue, M., Postsynaptic influences on the vestibular non-Deiters nuclei from primary vestibular nerve, Exp. Brain Res., 8 (1969) 190-200. 20 LaVail, J. H. and LaVail, M. M., The retrograde intraaxonal transport of horseradish peroxidase in the chick visual system. A light and electron microscopic study, J. comp. Neurol., 157 (1974) 303-358. 21 Lorente de N6, R., Anatomy of the eighth nerve, The Laryngoscope, XLIII, (1933) 1-38. 22 Maciewicz, R. J., Eagen, K., Kaneko, C. R. S. and Highstein, S. M., Vestibular and medullary brain stem afferents to the abducens nucleus in the cat, Brain Research, 123 (1977) 229-240. 23 Thomas, R. C. and Wilson, V. J., Marking single neurons by staining with intracellular recording microelectrodes, Science, 151 (1966)1538-1539.
31
VESTIBULAR NUCLEUS NEURONS RELAYING EXCITATION FROM THE ANTERIOR CANAL TO THE OCULOMOTOR NUCLEUS
M. YAMAMOTO, I. SHIMOYAMA and S. M. HIGHSTEIN University of Tokyo, Faculty of Medicine, Department of Physiology, Hongo, Bunkyo-Ku, Tokyo 176 (Japan) and (S.M.H.) Department of Neuroscience, Albert Einstein College of Medicine, Bronx, N. Y. 10461 (U.S.A.)
(Accepted September 22nd, 1977)
SUMMARY A morphophysiological approach was undertaken to determine which vestibular nucleus neurons relay excitation from the anterior canal to the IIIrd nucleus. In anesthetized rabbits HRP was iontophoresed into the IIIrd nucleus and cells filled with HRP reaction product (positive cells) searched for within the vestibular nuclear complex. By lesioning the MLF or brachium conjunctivum immediately after iontophoresis it was demonstrated that positive cells in the dorsum of the superior vestibular nucleus are backfilled via their axons which ascend in the brachium conjunctivum. By COntrast positive cells in the center of the superior nucleus are backfilled via their axons in the MLF. In electrophysiological experiments in the presence of a severed MLF the anterior canal was selectively stimulated' for orthodromic~ and the 3rd nucleus stimulated for antidromic, activation of vestibular nucleus neurons. Recording extracellularly with glass microelectrodes filled with fast green FCF the only ceils both orthoand 'antidromically activated were localized to the dorsum of the superior vestibular nucleus. It is concluded that cells dorsally located in the superior nucleus relay the disynaptic excitatory vestibulo-ocular reflex from the anterior canal to the contralateral 3rd nucleus via their axons which ascend in the brachium conjunctivum.
INTRODUCTION The purpose of this paper is t o determine which cells in the vestibular nuclei relay anterior canal excitation to the 3rd nucleus. There is physiological1~and morphologicalS, 7 evidence indicating that labyrinthine signals to oculomotor neurons are conveyed not only through the medial longitudinal fasciculus (MLF) but also via the brachium conjunctivum (BC). The presumed cells of the origin" of the BC pathway
32 to several subgroups of neurons in the third nucleus were located with a microstimulation technique to the region of the Y group of the vestibular nuclear complex and to the dentate nucleus of the cerebeUumL Both the Y group and the dentate nucleus receive primary vestibular afferent inputsa, 4, and it has been asserted that both project directly to the oculomotor nuclear complexS,7, 9. Anterior canal stimulation evokes reflex contraction in the ipsilateral superior rectus and the contralateral inferior oblique extraocular muscles in the rabbit 16. The BC pathway relays these anterior canal signals which remain after aspiration of the dentate nucleus 16. As it was suggested 9 that the dentate nucleus and the Y group contain the ceils of origin of the BC pathway it seemed logical to deduce that the anterior canal excitation is mediated by the Y group 16. However, Gacek 6 has suggested that direct afferent input to the Y group arises from the saccule and not from the utricle or semicircular canals. In contrast to Gacek's morphological results 6, Hwang and Poon ~a recorded from the Y group during stimulation of the saccule and found only multisynaptic activation of Y group neurons. In order to unravel this puzzle of which cells relay the anterior canal excitation to the 3rd nucleus a morphophysiological approach was undertaken. Our results show that there is a previously unnoticed population of ceils in the dorsal portions of the superior vestibular nucleus which project through the BC to the IIIrd nucleus. It is these superior nucleus cells which are now presumed to be the anterior canal relay neurons. MATERIALS AND METHODS
Morphological experiments Nine albino rabbits were anesthetized with intravenous urethane (400 mg/kg) and a-chloralose (40 mg/kg) and were mounted in a stereotaxic apparatus. Craniotomy was performed and the occipital lobe on the right side was aspirated to expose the surface of the superior colliculus. The whole 3rd nerve was stimulated stereotaxically 12. A glass microelectrode (2 M NaC1, resistance 2-4 Mf~) was inserted through the superior colliculus to identify the region of the 3rd cranial nucleus by observing the antidromic field potentials evoked by 3rd nerve stimulation. Once the limits of the nucleus were defined the recording electrode was replaced with a bevelled glass microelectrode filled with concentrated horseradish peroxidase (HRP) solution in Tris buffer pH 8.6. We attempted to place the H R P electrode 1 mm rostral to the caudal edge of the 3rd nucleus. H R P was iontophoresed using 1-3/~A 250 msec positive pulses every 500 msec for a total on time of 30-60 minS, 22. In 5 animals, immediately after the iontophoretic procedure the cerebellum was exposed and the vermal portions aspirated to expose the floor of the 4th ventricle. A transverse incision in the brain stem at a level several millimeters anterior to the superior vestibular nucleus was made. The intent of these lesions was to sever the median longitudinal fasciculus (MLF) and the pontine tegmentum bilaterally. As the pathways relaying disynaptic vestibuloocular signals have been demonstrated to be the M L F 7,1°,12,15, BC 5,1°,1g,16 and ascending tract of Deitersll the intent of our lesions was to interrupt
33 these pathways. Even though our lesions often impinged upon reticular structures we felt that this did not interfere with the purpose of this study, as reticular nuclei have not been implicated in the disynaptic vestibuloocular reflexes. In one rabbit the BC was interrupted by a transverse incision just posterior to the base of the inferior colliculus. In the remaining 3 animals the brain stem was left intact. Twenty to 24 h after the injection the animals were perfused with 10 ~ formalin in 0.1 M phosphate buffer at pH 7.4. The brains were removed from the skull and stored in cold phosphate buffer containing 30 ~ sucrose. On the following day serial frozen sections, 50/~m thick were cut, incubated for 30-60 min in hydrogen peroxide and 3,3'-diaminobenzidine tetrahydrochloride. The sections were then mounted on glass slides counter stained with 0.5~ cresyl violet, and drawn at 10 x magnification. Subsequently sections were examined with a light microscope and the locations of cells labelled with HRP reaction product were marked on the drawings.
Electrophysioiogical experiments Fifteen rabbits were used, anesthetized with chloralose-urethane. The trachea was cannulated, the rabbits were placed in a stereotaxic apparatus and artificially respired. Through a dorsal approach the bony labyrinth was visualized and the area overlying the ampulla of the anterior canal identified. A small hole was made in the bone overlying the ampulla and isolated stimulation of the anterior canal with fine bipolar wires was attempted 14. EMG electrodes were placed in the ipsilateral superior rectus and medial rectus extraocular muscles. The threshold for EMG activation of the superior rectus muscle via anterior canal stimulation was determined. Because the horizontal canal nerve is the most likely vestibular structure to be excited by current spread during anterior canal stimulation and it is known that the horizontal canal relays excitation to the medial rectus muscle, we determined how many multiples of threshold for superior rectus activation were necessary before medial rectus activation was observed. During the subsequent electrophysiological experiments, care was taken to insure that the stimulus intensity to the anterior canal was always below that which evoked excitation in the medial rectus muscle. Craniotomy was performed, the occipital cortex aspirated and a concentric bipolar stimulating electrode inserted through the superior colliculus to the region of the oculomotor nuclear complex. Final position of this stimulating electrode was adjusted by recording the field potentials following anterior canal stimulation. The cerebellum was exposed by craniotomy, its vermal portions aspirated and a broad deep cut of the pons rostral to the superior nucleus was made to interrupt the medial longitudinal fasciculus and the pontine tegmentum bilaterally. Single units in the area of the vestibular nuclei were isolated with a recording microelectrode filled with fast green FCF 2a during stimulation of the anterior canal for orthodromic and the oculomotor nucleus for antidromic activation. Sites of positive cells anti- and orthodromically activated were marked with fast green and were subsequently identified and indicated on histological reconstructions of brain stem sections.
34
A
3,d..c
H
~ - -
~ o
¢ 3,
D _~.~ ~ ~/~-.'.~> '~~
I
~/~
Fig. 1. H R P injection into the Iltrd nucleus with an intact brain stem. A - C illustrate the injection site. The black area indicates the region most heavily impregnated with reaction product, the shaded area is a zone of lighter, but still appreciable impregnation and the dotted lines indicate the limit of visible H R P diffusion. The most heavily impregnated area was limited to the right oculomotor nucleus. D - K illustrate the distribution of H R P positive cells in the brain stem. BC, brachium conjunctivum; DV, descending vestibular nucleus; DN, lateral cerebellar nucleus; LV, lateral vestibular nucleus; MV, medial vestibular nucleus; RB, restiform body; SV, superior vestibular nucleus; Y, vestibular Y group; III, oculomotor nucleus; VI, abducens nucleus.
35
Fig. 2. Light micrograph of HRP positive cells within the dorsal portions of the superior nucleus. Fine granules can be seen. The bar indicates 20 #m. RESULTS
(A) Morphology (1) Superior nucleus cells projecting through the brachium conjunctivum to the third nucleus (a) Control animals. The 3 intact animals had similar results; Fig. 1 illustrates a typical case. Nerve cells in and around the vestibular nuclei which contain horseradish peroxidase granular reaction product were regarded as positive cells (Fig. 2). Diffusely stained cells were discarded as they might have been artifactually stained due to interruption of their axons by the pipettes placed in the 3rd nucleus. In the case illustrated in Fig. 1, the dense HRP reaction product was mainly localized to the right oculomotor nuclear complex. The size of the injection site was defined as the limit of the spread of the homogeneous horseradish reaction product within the 3rd nucleus. Iontophoretic application of HRP in general yielded a restricted spread of label at the injection site. The site consisted of a central region which was densely stained with reaction product surrounded by a much more lightly stained zone. The limits of the site can be visualized in Fig. 1A, B and C. The dense black area within the oculomotor nuclear complex is the area of dense reaction product. The cross hatched area indicates the more lightly labeled zone and the dotted line indicates the limits of the visible reaction product. It is likely that only cells which project to the densely labeled portions of the injection site retrogradely transport horseradish peroxidase 22. In sections D through H, there are labeled cells in the superior nuclei bilaterally. In both the Y group (1I and J) and the superior nucleus positive cells are either spindle shaped with flattened nuclei or multipolar with round nuclei. The cells appear to lie in rows between the passing fibers. The size of these ceils is approximately 20-50 #m which is comparable to the diameters of cells within the center of the nucleus. Typical cells are illustrated in Fig. 2. In Fig. 1F the distribution of labeled cells within the bilateral superior nucleus is of interest. Contralateral to the injection site the labeled cells are concentrated within the dorsal parts of the superior nucleus, the central
36
A
B
Fig. 3. HRP injection into the IIIrd nucleus with a severed MLF and pontine tegmentum. Labeling as in Fig. 1. A indicates the injection site. B indicates the extent of the lesion of the brain stem (black area). In this case the injection was unilateral and only the contralateral superior nucleus contains positive cells.
portions of the nucleus being free of positive ceils. On the side ipsilateral to the injection site the dorsal portions of the nucleus are free of labeled cells, while the labeled ceils are concentrated in the central regions of the nucleus. The contralateral medial vestibular nucleus ( 1 H - K ) contains labeled ceils in its rostral portions but there are also labeled cells ipsilaterally (1 H). In addition the contralateral descending vestibular nucleus contains positive cells (1J and K). The Y group overlying the restiform body is clearly filled with many positive cells (1I and J). Additionally there are positive cells lying within the contralateral abducens nucleus (1I). These are presumably abducens internuclear neurons projecting to the contralateral medial rectus subgroup as has been described in the cat2,7,11. Cell bridges between the Y group and the dentate nucleus also contain positive cells. There were no positive cells in either the dentate nucleus of the cerebellum or in dorsal Deiter's nucleus. There were two additional
37
A
Fig. 4. HRP injection into the IIIrd nucleus with a severed BC. A shows the injection sites. B shows the extent of the BC lesion. Note that in this case the injection was bilateral and the left BC was cot. animals in the control group. In one animal the injection site covered the whole third nucleus bilaterally and the lightly stained area extended to the MLF. Positive cells in this case were found mainly in the medial and the superior vestibular nuclei. In the third case the injection site was small and localized within the caudal half of the bilateral third nuclei. Though the general features of the distribution of positive cells were similar to that in the first two cases the number of labeled cells were less abundant in the medial nucleus and descending nucleus. (b) Lesion experiments. In 5 rabbits horseradish peroxidase was iontophoresed into the 3rd nucleus unilaterally and the M L F was interrupted bilaterally. The extent of the lesion is exemplified in Fig. 3B. In all 5 of these cases positive cells
38
F
---
G
Fig. 5. Electrophysiologicai analysis of the distribution of neurons orthodromically activated by anterior canal stimulation and/or antidromically activated by llIrd nucleus stimulation. A: orthodromic activation of a neuron by anterior canal stimulation and B: antidromic activation of the same neuron by stimulation at the site marked in black in D. C: reproduction of Fig. IF. D and E: cross sections of the brain stem to indicate the site of stimulation within the IIIrd nucleus in D and the extent of the brain stem lesion in E. F - H : brain stem sections to indicate the location of cells orthodromically activated by anterior canal and antidromically activated by IIIrd nucleus stimulation (~k). Orthodromic only cells indicated by O; antidromic only cells indicated by 0 .
were seen in the c o n t r a l a t e r a l b u t n o t in the ipsilateral Y g r o u p ( 3 D - F ) . In a d d i t i o n some cells were f o u n d within the dorsal p o r t i o n s o f the c o n t r a l a t e r a l superior vestibular nucleus ventral to the BC a n d medial to the restiform b o d y (3C a n d D). I n a n o t h e r r a b b i t the BC was i n t e r r u p t e d unilaterally after the i o n t o p h o resis o f p e r o x i d a s e into the o c u l o m o t o r nucleus. In this case the i o n t o p h o r e t i c inject i o n e n c o m p a s s e d the c a u d a l two-thirds o f b o t h third nuclei (Fig. 4A). O n the side o f the intact BC cells in the Y g r o u p were labeled j u s t as described in the previous five p r e p a r a t i o n s ( 4 E - G ) . O n the lesion side however, there were no positive cells
39 within the Y group (Fig. 4). The distribution of labeled cells in the bilateral superior nuclei, in this case with a lesioned BC, complements the cases with M L F lesions. On the intact side positive cells are found throughout the superior nucleus (4C-E). That is, in its central portions as well as in the dorsal portions underlying the BC. However, on the lesion side only the central and caudal portions of the superior nucleus are stained (4C-E). The dorsal portions of the nucleus underlying the BC are not filled with positive cells. This can be seen quite clearly by comparing Fig. 4C and 3C. This implies that those positive cells underlying the BC are labeled via the BC, while those cells in the center of the nucleus are labeled via the M L F pathway. In another series of rabbits, both the BC and M L F were extensively interrupted on both sides. In this case, no positive cells were found anywhere within the vestibular nuclei. This control experiment eliminates the possibility that the above described results were an artifact of either endogenous peroxidases or of endogenous peroxidase uptake. ( B) Electrophysiological identification o f cells activated by anterior canal and oculomotor nuclear stimulation
After measuring the midline and the depth of the floor of the fourth ventricle with a microelectrode, systematic tracking was made throughout the area of the vestibular nuclear complex. Maps were drawn of the tracks indicating cells which were activated antidromically by third nucleus or orthodromicaUy by anterior canal stimulation. Fig. 5A and B illustrate the extracellular records taken from a unit which was both orthodromically (5A) and antidromically (5B) activated. Orthodromic activation times were typically 1-1.5 msec. There was some jitter in the latency to activation as expected with orthodromic activation. It has been reported 19 that superior nucleus neurons can be activated in as little as 0.8 msec following supramaximal eighth nerve stimulation. However, in the present experiment stimulation intensities employed were rather weak in order to insure isolated activation of the anterior canal. Thus the latencies to activation (1-1.5 msec) probably lie within the monosynaptic range. Antidromic activation was all-or-none with fixed latency. The typical latency for antidromic activation was TABLE I The distribution of electrophysiologically identified cells in the rabbit's brain stem
Abbreviations: AC, anterior canal; A and O, antidromically and orthodromically activated cells; A, Cells activated only antidromically; O, Cells activated only orthodromically; LV, Deiter's nucleus; DN, dentate nucleus of the cerebellum; SV, superior vestibular nucleus; VLD, ventrolateral vestibular nucleus; NVC, nucleus cerebello-vestibulosis; BC, brachium conjunctivum; SV-Y, area between the Y group and SV; SV-BC, area between the BC and the SV; Y, Y group. The numbers indicate the number of cells identified as lying within the structures indicated in each column. AC
O-}-A A O
LV
23
DN
SV
VLD
NVC
BC
SV- Y
SV-BC
17 14 52
1 2
2
4
5 19 12
2
2
Y
7 1
40 0.5-0.6 msec. When double shock stimulation was given to a unit which was antidromically activated, the latency to the activation following the second stimulus was often significantly longer than the first (not illustrated). This is presumptive evidence for antidromic activation. Cells were counted as being activated by the anterior canal if they could be activated with low stimulus intensity as mentioned in the Methods. The electrophysiological results herein reported are based upon 16l units. Many other units were discarded as they could not be accurately located on the histological reconstructions. Fig. 5 illustrates the locations of units from 4 experiments. It can be seen (5F-H) that units which are ortho- and antidromically activated (~r) tend to lie in the dorsal parts of the superior nucleus, while units which are only orthodromically activated (O) lie either among the incoming fibers of the eighth nerve or in the more ventral portions of the superior nucleus. Records of the two units anti- and orthodromically activated within the BC (5F) were presumably taken from the dendrites of superior nucleus cells. Several units only antidromicaUy activated lie within the lateral nucleus of the cerebellum or adjacent to it in the interpositus. Fig. 1F is reproduced in 5C for comparison with the electrophysiological results (see Discussion). The distribution of 161 cells which were ortho- and/or antidromically activated is illustrated in Table I. DISCUSSION The present results support the previous electrophysiological evidence of a direct vestibular oculomotor projection through the BC. The rostral dorsal portions of the superior nucleus were demonstrated electrophysiologically to contain cells which carry signals from the anterior canal relaying them through the BC to the contralateral third nucleus. Morphological experiments support this view. It is striking to note the distribution of labeled cells in Fig. 1F (and 5C) within the bilateral superior nuclei. In this case the injection site is limited to the right oculomotor nucleus. Results herein presented in Figs. 3-5 suggest that the dorsally located cells in the left (contralateral to the injection) superior nucleus were labeled with HRP retrogradely transported via their axons in the BC while the centrally located cells in the right (ipsilateral) superior nucleus were labeled via their M L F projection. This view gains credence because the right third nucleus contains the motoneurons innervating the right inferior oblique and the left superior rectus 9, both of which are activated by left anterior canal stimulation 16. In this context it is also noted that this identical distribution of labeled neurons has been demonstrated in the cat by Graybiel and Hartweig 7 in their Fig. 1. That is, in Graybiel's paper if the injection of peroxidase is limited to one oculomotor nuclear complex (as in Fig. 1), then the identical distribution of cells in the bilateral superior nuclei as described for the rabbit is observed. However, when the injection encompassed both third nuclei (Graybiel and Hartweig, Fig. 2) the distinctive distribution of cells in the superior nuclei is lost. It can now be suggested that the oculomotor output of the superior nucleus is divided into at least two portions. The portion containing cells which lie directly underneath the BC is apparently relaying excitatory signals through the brachium
41 to the contralateral oculomotor nuclear complex. The superior nucleus cells relaying inhibition up the ipsilateral M L F 1°,12,15 lie within the central portions of the nucleus. It was previously thought that the cells in the periphery of the superior nucleus did not receive any input from the labyrinth 4. Our data suggests that the dorsal portions of superior nucleus receive primary input from the anterior canal. Gacek 6 also supports the view that the entire superior nucleus receives afferents from the labyrinth. Further our results (Fig. 2) indicate that neurons dorsally located within the nucleus can be as large as those in the center. Gacek (personal communication) also finds large ceils dorsally located. Secondly, flocculus Purkinje ceils were shown to end predominantly in the center of the superior nucleus 1. Since the flocculus inhibits the excitatory pathway from the anterior canal xT, we infer that it must also terminate on the dorsum of the superior nucleus. Alley (personal communication) substantiates this view utilizing labeled amino acids to map flocculus projections. The area of the Y group contains many H R P positive ceils after third nucleus injection as do the cell bridges between the Y group and dentate nucleus of the cerebellum. This area of cell bridges appears identical to the area which Lorente de N6 ~1 originally called the nucleus cerebellovestibulosis in the guinea pig. Gacek (personal communication) also confirms that the primary afferents from the saccule project to the area of these cell bridges. Y group cells were surveyed with a microelectrode but were not activated by isolated stimulation of the anterior canal. (Ito et al., unpublished; Yamamoto et al., this paper, Table I). Evidence to date6,13, 21 supports the idea that the Y-group is the relay nucleus for the sacculoocular reflex. Ceils projecting to the third nucleus were not found in the dentate nucleus of the cerebellum. Graybiel et al. 7 also failed to define labeled cells within the dentate except for a very few. Whether the failure of labeling in cats and rabbits indicates the absence of dentate ceils projecting to the third nucleus is still open to question. Experiments on this point are currently in progress. Lastly abducens internuclear neurons demonstrated in the cat 2,11,7 are found in the present study thus confirming this important pathway in the rabbit, and suggesting some common features in the organization of horizontal as well as vertical gaze. ACKNOWLEDGEMENTS The authors would like to thank Prof. M. Ito for his critical reading of the manuscript and for his support during this study. This work was supported by Grants No: N I H 5 K04 EY 00003-02, N I H 5 R01 EY 01670-02 and NSF OIP 74-13621.
REFERENCES 1 Angaut, P. and Brodal, A., The projection of the 'vestibulo-cerebellum' onto the vestibular nuclei in the cat, Arch. itaL BioL, 105 (1965) 441-479. 2 Baker, R. and Highstein, S. M., Physiological identification of interneurons and motoneurons in the abducens nucleus, Brain Research, 91 (1975) 292-298.
42 3 Brodal, A. and Hoivik, B., Site and mode of termination of primary vestibulocerebellar fibers in the cat, Arch. ital. Biol., 102 (1964) 1-21. 4 Brodal, A. and Pompeiano, O., The vestibular nuclei in the cat, J. Anat. (Lond.), 91 (1957) 438454. 5 Carpenter, M. B. and Strominger, N. C., Cerebello-oculomotor fibers in the rhesus monkey, J. comp. Neurol., 123 (1964) 211 230. 6 Gacek, R., The course and central termination of first order neurons supplying vestibular end organs in the cat, Acta Otolaryng. (Stockh.), Suppl., 254 (1969) 1-66. 7 Graybiel, A. M. and Hartwieg, E. A., Some afferent connections of the oculomotor complex in the cat: an experimental study with tracer techniques, Brain Research, 81 (1974) 543-551. 8 Graybiel, A. M. and Devor, M., A microelectrophoretic delivery technique for use with horseradish peroxidase, Brain Research, 68 (1974) 167-173. 9 Highstein, S. M., The organization of the vestibulo-oculomotor and trochlear reflex pathways in the rabbit, Exp. Brain Res., 17 (1973) 285-300. 10 Highstein, S. M., Organization of the inhibitory and excitatory vestibulo-ocular reflex pathways to the third and fourth nuclei in rabbit, Brain Research, 32 (1971) 218-224. 11 Highstein, S. M. and Baker, R., Termination of internuclear neurons of the abducens nuclei on medial rectus motoneurons, Neuroscience Abs., 2 (1976) 398. 12 Highstein, S. M., lto, M. and Tsuchiya, T., Synaptic linkage in the vestibulo-ocular reflex pathway of rabbit, Exp. Brain Res., 13 (1971) 306-326. 13 Hwang, J. C. and Poon, W. F., An electrophysiological study of the sacculo-ocular pathways in cats, Jap. J. Physiol., 25 (1975) 241-251. 14 Ito, M., Nisimaru, N. and Yamamoto, M., The neural pathways mediating reflex contraction of extraocular muscles during semicircular canal stimulation in rabbits, Brain Research, 55 (1973) 183-188. 15 lto, M., Nisimaru, N. and Yamamoto, M., The neural pathways relaying reflex inhibition from semicircular canals to extraocular muscles of rabbits, Brain Research, 55 (1973) 189-193. 16 Ito, M., Nisimaru, N. and Yamamoto, M., Pathways for the vestibulo-ocular reflex excitation arising from semicircular canals of rabbits, Exp. Brain Res., 24 (1976) 257-271. 17 Ito, M., Nisimaru, N. and Yamamoto, M., Specific patterns of neuronal connections involved in the control of rabbits' vestibulo-ocular reflexes by the cerebellar flocculus. J. Physiol. (Lond.), 265 (1977) 833-854. 18 lto, M., Nisimaru, N. and Yamamoto, M., Unpublished. 19 Kawai, N., Ito, M. and Nozue, M., Postsynaptic influences on the vestibular non-Deiters nuclei from primary vestibular nerve, Exp. Brain Res., 8 (1969) 190-200. 20 LaVail, J. H. and LaVail, M. M., The retrograde intraaxonal transport of horseradish peroxidase in the chick visual system. A light and electron microscopic study, J. comp. Neurol., 157 (1974) 303-358. 21 Lorente de N6, R., Anatomy of the eighth nerve, The Laryngoscope, XLIII, (1933) 1-38. 22 Maciewicz, R. J., Eagen, K., Kaneko, C. R. S. and Highstein, S. M., Vestibular and medullary brain stem afferents to the abducens nucleus in the cat, Brain Research, 123 (1977) 229-240. 23 Thomas, R. C. and Wilson, V. J., Marking single neurons by staining with intracellular recording microelectrodes, Science, 151 (1966)1538-1539.
