Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Vestibular and Somatic Inputs to Cells of the Lateral and Medial Vestibular Nuclei of the Cat1 VICTOR J. WILSON The Rockefeller University SUMMARY An important input to Deiters' cells is that from the labyrinth, which includes fibers from static receptors. The input from the labyrinth is seen in a larger fraction of the cells projecting to the cervical and thoracic cord than of the cells projecting to the lumbosacral cord. Both groups of cells can be facilitated by impulses ascending the spinal cord. These impulses are due to activity in a variety of peripheral nerves coming from different receptors, but apparently not from primary spindle endings or Golgi tendon organs. The cells of Deiters' nucleus, that influence the excitability of mo- toneurons at all levels of the spinal cord via the vestibulospinal tract, are therefore themselves impinged upon by a variety of inputs that share in the regulation of their excitability. Some cells in the medial vestibular nucleus project to the spinal cord, but many more project rostrally. Both types of projecting cells are almost completely absent from the caudal region of the nucleus, as is the monosynaptic input from primary vestibular fibers. The vestibular input originate? in the horizontal canal and utricle, and probably in other parts of the labyrinth. Electric stimulation of the labyrinth activates many projecting cells as well as many cells without long axons. There is also a somatic input to cells in the medial nucleus, and it is of particular interest that vestibular and somatic inputs, as well as commissural inputs and inputs from fibers descending from higher levels of the central nervous system, converge on many cells lacking long axons projecting rostrally or to the spinal cord. It is probable that among these cells there are interneurons that regulate the activity of projecting cells. INTRODUCTION In recent years a remarkable amount of detailed anatomical information has become available about the organization of the vestibular nuclei, the projection of their cells and the inputs to them, to a large extent through the investiga- tions of Brodal and his collaborators in Oslo. It is profitable to attempt to relate this anatomical knowledge to the properties of neurons in the vestibular nuclei, as revealed by electrophysio- logical methods. Such studies have been under- taken in several laboratories, including ours. So far, we have concentrated our efforts on an 1 Work in the author's laboratory was supported in part by grant 5R01 NB 02619 from the National Institute of Neurological Diseases and Blindness, I'SPHS. analysis of the organization of the lateral and medial vestibular nuclei of the cat, and of vestibular and somatic inputs to cells in these nuclei. All our experiments have emphasized study of cells identified by location and projec- tion, since only in this way is it possible to relate physiological results to the anatomical data at our disposal. In this paper I will describe some of the results of experiments that have been presented in greater detail elsewhere (refs. 1 to 4). All our experiments have been performed on acutely decerebellated cats, anesthetized by intraperitoneal injections of chloralose (40 mg/kg) and urethane (800 mg/kg) dissolved in poly- ethylene glycol, paralyzed by gallamine tri- ethiodide (Flaxedil, American Cyanamid Co.), and artificially respired. A diagrammatic repre- sentation of the experimental arrangement is
146 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION shown in figure 1. Recording of the activity of single neurons in the vestibular nuclei was extracellular, by means of glass micropipets filled with 2 M NaCl saturated with Fast Green FCF. With such electrodes it is possible to eject a small amount of dye (refs. 5 and 6), which was frequently done at the bottom of electrode tracks. The dye marks were easily found in serial histological sections of the vestibular region prepared after each experiment, which made accurate localization of the tip of the recording electrode possible. During experiments, cells were sometimes found by the presence of spontaneous or synaptically evoked activity, more often by antidromic activation. Anti- dromic invasion of Deiters' cells resulted from stimulation of the vestibulospinal tract at various levels of the spinal cord; cells in the medial nucleus were activated antidromically either by DEITERS' EXPERIMENTS RECORDING ELECTRODE MEDIAL NUCLEUS EXPERIMENTS CONTRALATERAL LABYRINTH IPSILATERAL LABYRINTH FIGURE I. âDiagram of experimental arrangement. A, ex- periments on Deiters' nucleus. Stimulating electrodes (S) at C3âC4 and L3âL4 were used for antidromic activation of vestibular neurons. Cells activated by the lumbar and cervical electrodes were classified as L-cells, while cells stimulated only by the cervical electrodes were classified as C-cells. The other stimulating electrodes were used to stimulate various limb nerves, and the ipsilateral vestibular nerve. B, experiments on the medial nucleus. Electrodes for antidromic stimulation were placed in the medial longi- tudinal fasciculus rostral to Deiters' nucleus (MLF) and in the upper cervical cord (DMLF). an electrode inserted near the descending medial longitudinal fasciculus at the level of the third or first cervical segment, or by an electrode placed in the medial longitudinal fasciculus about 2 mm rostral to Deiters' nucleus. Ves- tibular afferent fibers were activated by electrical stimulation by means of an electrode inserted in the scala vestibuli. In experiments on Deiters' nucleus, only the ipsilateral labyrinth was stimu- lated, but contralateral stimulation was added in many medial nucleus experiments. In one series of experiments, natural stimulation of static receptors was achieved by tilting (ref. 7). Somatic afferent fibers were activated by electrical stimu- lation of various muscle, cutaneous, and mixed limb nerves, as well as by stimulation of spinal tracts with the same electrodes used for anti- dromic stimulation of cells in the vestibular nuclei. Further details of experimental proce- dures can be found in the original publications (refs. 1 to 3). THE LATERAL VESTIBULAR NUCLEUS Many cells in this nucleus project to the spinal cord in the vestibulospinal tract, ending at all levels from upper cervical to sacral (refs. 8 to 10). Electrical stimulation of Deiters' nucleus results in facilitation, monosynaptic or polysynaptic, of extensor motoneurons at all these levels of the spinal cord (refs. 11 and 12). Anatomical, as well as some physiological, investigations have suggested that the nucleus is somatotopically organized; specifically, cells projecting to the cervical cord are to be found predominantly in the rostroventral part of the nucleus while cells projecting to the lumbosacral cord are mainly located in its dorsocaudal region (ref. 13). To a certain extent this distribution has been confirmed by recent electrophysiological studies (refs. 2 and 14). Our experiments, however, show that there is considerable departure from the ideal somatotopic pattern. Localization of cells found in tracks marked by dye ejection in- dicates that the dorsocaudal part of the nucleus does contain mainly cells whose axons extend to the lumbosacral cord (henceforth designated L-cells). In the rest of the nucleus, L-cells and C-cells (the latter projecting to the fore- limb region and thoracic cord) are intermingled,
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAR NUCLEI 147 C-cells outnumbering L-cells by a relatively small margin (ref. 2). Such serious blurring of somatotopic organization must be taken into account when considering the functional meaning of the termination of different types of afferents in specific regions of the lateral vestibular nucleus. It has been suggested previously that Deiters' nucleus can be divided into parts that are not equivalent functionally. Our findings on the distribution of spontaneous activity, which is much more prevalent dorsally than ventrally in the type of preparation we have used (ref. 2), and on the distribution of the vestibular input, to- gether with the results of others on the location of cells inhibited by stimulation of the cerebellar cortex (ref. 15), indicate that the most mean- ingful dividing line is between the dorsal and ventral regions of the nucleus. As we shall see below, the ventral part of the nucleus re- ceives a pronounced input from the labyrinth. The dorsal part, on the other hand, receives a strong inhibitory input from the cerebellar cortex. It is therefore reasonable to consider ventral Dieters' cells as relays between labyrinth and spinal cord, while dorsal Deiters' cells are part of the efferent system of the cerebellar cortex. Vestibular Input to Cells in the Lateral Nucleus It was shown by Walberg, Bowsher, and Brodal (ref. 16) that vestibular afferents terminate principally in the rostroventral part of Deiters' nucleus. This has recently been confirmed by Mugnaini, Walberg, and Brodal (ref. 17) who showed, in addition, that vestibular afferents terminate not only on small- and medium-size cells, but also on some giant neurons. In ex- cellent agreement with these findings, Ito and his colleagues (ref. 15) have observed monosynaptic excitatory postsynaptic potentials (EPSP's) in Deiters' neurons on stimulation of the vestibular nerve, and our results (ref. 2) show that such monosynaptic excitation, illustrated in figure 2, is found almost exclusively in the ventral part of the nucleus. Among the cells that we sampled, 31 of 61 C-cells (51 percent) and 18 of 80 L-cells (22 percent) could be fired monosynaptically by FIGURE 2.âEffect of stimulation of the labyrinth on a rostral L-cell in Deiters' nucleus. Each picture consists of several superimposed sweeps, recorded extracellularly. A: I/sec. The cell is driven monosynaptically from the labyrinth and responds to every shock; there is a large P-wave, represent- ing the incoming afferent volley, right after the stimulus arti- fact. B: 5/sec. Cell misses occasionally, revealing the underlying NI potential. This is a postsynaptic (monosyn- aptic) field potential. C: 20/sec. Cell no longer responds, but a sizable Nl remains. Time mark, msec; voltage cali- bration, 500 IJ.V. Negative deflection upward in this and succeeding figures. (From ref. 3.) stimulating the labyrinth with a single electric shock. Some of these cells had axons the con- duction velocity of which exceeded 100 m/sec; these were giant cells. As shown in table 1, almost all cells that fired monosynaptically were- located in the ventral part of the nucleus. In addition to cells fired monosynaptically, a few (16/141, or 11 percent) were fired polysynaptically with a longer latency. This relatively small number of cells fired polysynaptically is in con- trast to the much greater number of medial nucleus cells so fired (ref. 4, and below). TABLE I.âFraction of Cells Driven Monasyn- aptically by the Labyrinth in Different Part of Deiters' Nucleus Rostral Middle Caudal Dorsal 1/3C 0/8 0/4C 1/7L 0/10L Ventral 6/12C 3/6 6/9C 9/1 1L 2/6L This table is based on 76 cells found in tracks containing a dye mark. Location of cells was determined from an analysis of serial frozen sections cut at 40 microns and stained with thionin. L, cell projecting to lumbosacral cord; C, cell projecting to cervicothoracic cord. Data from ref. 2 and from unpublished observations by B. W. Peterson.
148 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION Although stimulation of the labyrinth excites more C-cells than L-cells, in keeping with previous anatomical and physiological findings (ref. 13), the number of L-cells excited is far from negligible. A relatively direct pathway therefore exists between the labyrinth and segmental motor mechanisms at all levels of the spinal cord. The question may be asked, What part or parts of the labyrinth give rise to the afferent fibers that reach the lateral vestibular nucleus? In 1933, Lorente de No (ref. 18; see also ref. 19) showed that while there is overlap between the projections of different parts of the labyrinth within the vestibular nuclei, there is a clear projection to Deiters' nucleus from the utricle, the receptor for position sense. Adrian (ref. 20) found that vestibular neurons in the cat were affected by tilt, and subsequently Duensing and Schaefer (ref. 21) studied this effect of tilt on various cells in the vestibular nuclei, some of them located in Deiters' nucleus. Recent detailed investigations by Peterson (ref. 7) have shown that the spontaneous activity of many lateral vestibular cells, including cells with axons in the vestibulo- spinal tract, can be increased or decreased by tilting (see also ref. 22). In agreement with the findings described above, Peterson observed that cells strongly affected by tilting were generally found in the ventral part of Deiters' nucleus; electrical stimulation revealed that most of the strongly affected cells received a monosynaptic input from the labyrinth. This monosynaptic input could come from various receptors, as there is evidence that canal and utricular receptors often converge on cells in the lateral nucleus, as well as in other nuclei (refs. 7, 21, and 23). Somatic Input to Cells in the Lateral Vestibular Nucleus It has been known for some time that fibers in the restiform body may give off collaterals to the vestibular nuclei (cf. ref. 24). These fibers may include dorsal spinocerebellar tract fibers, but at least some of the spinovestibular fibers that have been described recently (refs. 25 and 26) are not collaterals of the dorsal spinocere- bellar tract. Spinovestibular fibers are scanty in numbers and terminate, among other places, in the caudal part of Deiters' nucleus. In accord with these observations, short-latency excitatory postsynaptic potentials have been observed in some Deiters' cells following stimulation of the ventral and lateral funiculi of the cervical spinal cord (ref. 14). What the functional meaning of this pathway is remains to be determined. It is unlikely to play a very important role in regulating the activity of Deiters' cells; in our experiments, stimulation of the spinal cord or of peripheral nerves caused almost no short-latency modifica- tion of the activity of Deiters' cells (refs. 1 and 2). We have observed that stimulation of various forelimb and hindlimb nerves can produce a relatively long-lasting facilitation of the firing of single units in Deiters' nucleus, as is shown in figure 3 (refs. 1 and 2); not unexpectedly, facilita- tion can also be evoked by stimulation of muscle and cutaneous afferents from the neck (un- published observations by R. M. Wylie, M. Yoshida, and V. J. Wilson). Unlike stimulation of the labyrinth that could fire silent cells, stimula- tion of peripheral nerves usually was able only to facilitate cells with ongoing spontaneous activ- ity, of which there were many. Some other characteristics of this peripheral activation of Deiters' cells are as follows (refs. 1 and 2): (1) It has a latency near 20 msec (on stimulation of hindlimb nerves), and the pathway that pro- duces it has a spinal conduction velocity ranging from 10 to 80 m/sec, usually 40 m/sec or less; (2) the facilitation lasts 100 to 200 msec; (3) while FIGURE 3.âActivation of a Deiters' L-cell by peripheral stimulation. Extracellularly recorded spikes are displayed on the lower beam while the afferent spike and cord po- tential, evoked by stimulation of hindlimb nerves and re- corded at the cord-dorsal root junction, are displayed on the upper beam. A, spontaneous firing; B, effect of a strong shock to many ipsilateral nerves simultaneously; C, effect of a stimulus to the superficial peroneal nerve, strong enough to activate many delta fibers. Calibrations: time, 10 msec: amplitude for lower beam, 500 p.V. Spikes retouched. (From ref. I.}
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAR NUCLEI 149 single shocks to cutaneous or mixed nerves are effective in producing facilitation, multiple shocks to muscle nerves are usually required. An important result is that stimulation of ipsi- lateral and contralateral forelimb and hindlimb nerves often produces facilitation of the same Deiters' cells, and these may be cells projecting to the forelimb or hindlimb regions of the spinal cord. The facilitation is not organized somato- topically (ref. 2). Similar convergence was ob- served previously (refs. 27 and 28), but these earlier experiments were performed in animals with the cerebellum intact. Because of the localized distribution of spinovestibular fibers, the extensive convergence was ascribed to passage of impulses through the cerebellum. Our experiments on decerebellate animals show that this convergence is characteristic of the spinal or bulbar components of the ascending pathway. The nature of the ascending pathway is apparently quite complex and it may include pathways through the reticular formation as well as collaterals of spino-olivary, spinocerebel- lar, and reticulocerebellar fibers (refs. 1, 15, and 29). Finally, in our experiments, just as in those of Pompeiano and his colleagues (refs. 27 and 28), the change in excitability of Deiters' cells due to peripheral stimula- tion was usually facilitatory and only infrequently inhibitory. Impulses reaching the cerebellar cortex could activate Purkinje cells, which are inhibitory cells (ref. 30) and which project directly onto dorsal Deiters' neurons. Of course, impulses relayed through the cerebellar cortex and deep nuclei can produce not only inhibition but also facilitation and disinhibition of Deiters' cells (ref. 15), and the net effect resulting from different peripheral stimuli remains to be determined. Nevertheless, it would be expected that stimulation of peripheral nerves in animals with the cerebellum intact should produce more inhibition of Deiters' cells than has so far been described, and systematic investigation of this matter seems to be required. What kind of afferent fibers, when stimulated, produce facilitation of cells in the lateral vestib- ular nucleus? This question has been asked in experiments performed in animals with the cere- bellum intact as well as in decerebellate animals, and there is considerable agreement among the results obtained (refs. 1,2, and 28). Facilitation results when cutaneous or mixed forelimb and hindlimb nerves are stimulated with very weak shocks, and increases as the shocks are strength- ened to 20 or more times the threshold of the largest fibers in the nerve; impulses in large and small fibers can lead to facilitation of Deiters' cells. Evidently the large fibers that are effec- tive in producing facilitation and that are found in mixed nerves are not muscle afferent fibers. Stimulation of hindlimb muscle nerves does not result in facilitation until the shocks are strong enough to stimulate fibers larger than group I, whether the cerebellum is in place or removed (fig. 4; see refs. 1, 2, and 28). The same is true for stimulation of forelimb muscle nerves in decerebellate animals (fig. 4; ref. 2); in animals with an intact cerebellum it is possible that stimulation of the smaller forelimb group I fibers may influence Deiters' cells (ref. 28), but the data presented so far are not sufficient for critical evaluation of this possibility. It is interesting that cells in the brainstem reticular formation also are not affected by stimulation of hindlimb, or forelimb, group I fibers (ref. 31). Group I fibers activate some types of spinocerebellar fibers (ref. 32), and stimulation of group I fibers from some hindlimb muscles, particularly quadriceps, evokes field potentials in the olive (ref. 33). As it is likely that collaterals of some spinocere- bellar or olivocerebellar fibers terminate in Deiters' nucleus (see above), the possibility remains that searching by means of intracellular recording, or by means of extracellular excita- bility testing more exhaustive than so far em- ployed might reveal, even in decerebellate ani- mals, group la and Ib inputs to the cells of Deiters' nucleus that have so far escaped detec- tion. This possibility is even stronger in animals with the cerebellum intact. A group la input from hindlimb nerves (gastrocnemius-soleus), activating Deiters' cells by disinhibition via the cerebellum, has recently been assumed in ex- plaining long-latency facilitatory effects on H reflexes in human subjects (ref. 34). Facilitation of Deiters' neurons can be pro- duced by weaker stimuli when stimulating fore- limb muscle nerves than when stimulating hind-
150 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION limb muscle nerves, as shown in figure 4. It is generally accepted that effects that first appear when the stimulus to a muscle nerve reaches 5r A. Gastrocnemius-Soleus and Hamstring o. I Â° Â°- T- I 1.5- l.5T 2 2- |2.5-!3-l5.5-4- 2.5 1 3 BsUI 5 5- 10 10- 12- B. Anconeus and Lateral Head of Triceps O 4 T- LST l.S-2-.5-s / 2 / -!s-k5-4- /3.!|Â«l 5 6- 10- 12- 25 Multiple of group I threshold FIGURE 4. â Comparison of facilitating action on Deiters' cells of a triple shock (300/sec) to hindlimb (A) and fore- limb (B) muscle nerves. Results of stimulation of two dif- ferent hindlimb nerves were merged to construct graph \; of two forelimb nerves to construct graph B. Results from 12 cells were used for \,from 11 cells for B; three cells were common to both populations. T is the threshold of the largest fibers in the nerve; i.e., group I threshold. The bars show the standard deviation of each mean; n ranged from 5 to 50, and was usually 10 or more. For 20 of the cells, measurements were made from 100-msec sweeps; 200-msec sweeps were used for the other three cells. (From ref. 2.) approximately two times the threshold of the largest fibers in the nerve are due to activation of group II fibers, while effects starting at 5 to 10 times threshold are due to activation of group III fibers. From this and figure 4 it is obvious that, in decerebellated cats, hindlimb fibers in the group III category must be stimulated for facili- tation to appear, but that it is sufficient to stimu- late group II fibers in the forelimbs to evoke facilitation. Apparently hindlimb group II fibers are effective in animals with an intact cerebellum (ref. 28), but this, as well as any possi- ble contribution of group I fibers in such prepara- tions, needs further investigation. We have, therefore, a reasonable amount of information about the size of muscular and extra- muscular afferent fibers that, when stimulated, can influence the excitability of Deiters' cells. There is also some information about the recep- tors these fibers supply. Most interesting is the fact, indicated by the ineffectiveness of electrical stimulation of group I fibers, that activation of the primary endings of muscle spindles, or of tendon organs, has little or no influence on cells of the lateral vestibular nucleus. Apparently the vestib- ulospinal system, closely involved in the regula- tion of posture and muscle tone, is little affected by information originating in the most important muscle receptors. Since group II fibers in muscle nerves do not necessarily innervate only the secondary endings of the spindles (ref. 35), and since some muscle nerves may be contami- nated by joint afferents (cf. ref. 31), it is difficult to estimate the contribution of fibers from spindle secondary endings to the facilitation produced by stimulation of group II fibers. As for other receptors, Pompeiano and Cotti (ref. 27) found that Deiters' units could be facili- tated by manipulation of the limbs, tendon tap- ping, hair movement, and even by stimuli to the snout. In the experiments of Frederickson, Schwarz, and Kornhuber (ref. 36), deep somatic stimuli, particularly joint movement, were much more effective than exteroceptive stimuli in exciting vestibular cells. Most of the cells they studied, however, were located in the medial and descending nuclei, and a detailed investigation of the effect of different types of somatic stimuIi on Deiters' cells Still needs to be done.
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAR NUCLEI 151 Convergence of Vestibular and Somatic Inputs Vestibular and somatic imputs converge on many Deiters' cells. For example, in our ex- periments (ref. 2) most of the cells that were driven by stimulation of the labyrinth, mono- synaptically or polysynaptically, and that were spontaneously active (making detection of an ascending facilitatory input possible), were also facilitated by stimulation of leg nerves. While most of the facilitated cells studied in these experiments were located in the dorsal part of the nucleus, subsequent experiments have shown that ventrally placed cells can also be facilitated by impulses originating in the spinal cord, and these same experiments have shown that somatic and vestibular convergence takes place in cells whose activity is modified by static position changes (unpublished observations by B. W. Peterson). Convergence of vestibular and nonvestibular synaptic inputs (nonvestibular inputs include those transmitted via the cere- bellum as well as direct ones) has also been observed by Ito and'his collaborators"(ref. 15). It should be noted that convergence is seen in cells whose axons project to many levels of the spinal cord in the vestibulospinal tract (ref. 2). Cells projecting to the level of neck motoneurons, which are particularly influenced by vestibular inputs (refs. 37 and 38) and by stimulation of Deiters' nucleus (ref. 12), have not yet been studied. THE MEDIAL VESTIBULAR NUCLEUS The axons of many medial nucleus cells project rostrally, mainly in the medial longi- tudinal fasciculus (MLF), and terminate predom- inantly in the extraocular motor nuclei (refs. 39 and 40). Other cells send their axons to the spinal cord in the descending medial longi- tudinal fasciculus (DMLF; refs. 8, 40, and 41), as do some cells in the descending vestibular nucleus (ref. 42). The fibers to the spinal cord appear to be small in number, do not de- scend below midthoracic levels, and terminate on cells of laminae 7 and 8 (refs. 10 and 41). There are in the medial nucleus not only cells with long axons, but also a significant number of cells with axons that arborize within the vestibu- lar nuclei (refs. 13 and 43). This is in contrast to Deiters' nucleus; most of its cells are be- lieved to have long axons (ref. 13). I will refer to medial nucleus cells without long ascending or descending axons as interneurons, and will include among them commissural cells project- ing to the contralateral vestibular nuclei (ref. 44). In discussing our work on the medial nucleus (refs. 3 and 4), I will limit myself to describ- ing some vestibular and somatic inputs to the different types of cells in this nucleus. First, however, it is necessary to consider some as- pects of the nucleus' organization. For convenience, we have divided the nucleus into 10 areas equal in length. These areas are illustrated in figure 5; 1 is the most caudal and 10 the most rostral. The landmarks in these areas remained constant from one animal to another, making it possible to pool the data for cell location obtained in all experiments. Our findings on the relative number of units pro- jecting rostrally (MLF cells) and caudally (DMLF cells), and-on the location-ef-these units, FIGURE 5. â Distribution in areas I to 9 of the medial vestibu- lar nucleus of cells that were driven (filled circles) and were not driven (open circles) monosynaptically by stimulation of the labyrinth at a rate of 1/sec. Area 1 is the most caudal tenth of the nucleus; area 9 is very rostral. Abbreviations: D, descending vestibular nucleus; DV, dorsal nucleus of the vagus; G, genu of the facial nerve; L, lateral vestibular nucleus of Deiters'; N VII, facial nerve. NST, nucleus of the solitary tract; PH, nucleus prepositus hypoglossi; SA, stria acustica; ST. solitary tract; VI, nucleus of the abdu- cens nerve. (From ref. 4.)
152 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION are shown in table 2, and can be summarized as follows: (1) 41 percent of medial nucleus cells tested projected rostrally in the MLF, while only 17 percent projected to the spinal cord in the DMLF. This is in agreement with the results of anatomical investigations, that also show the rostrally projecting axons greatly to outnumber caudally directed axons. (2) Contrary to our expectations, very few cells had long axons pro- jecting both rostrally and caudally. Even though the number of dichotomizing cells re- vealed in our experiments is probably an under- estimate, there are apparently fewer such cells than previously believed (ref. 13). (3) Most cells projecting rostrally and caudally were lo- cated in areas 4 to 10; i.e., in the rostral two- thirds of the nucleus. (4) A substantial number of cells seemed to lack long axons. Some of these cells undoubtedly had long axons that were not excited by our stimuli, either because their threshold was too high or because they projected to regions we did not stimulate (e.g., the cerebellum, ref. 13), but others were cer- tainly interneurons. TABLE 2.-Fraction of Units Tested Projecting Into MLF AND DMLF Area Projection into MLF Projection into DMLF Nonprojecting units 1 1/6 1/6 4/6 2 0/21 3/22 14/18 3 3/33 0/20 19/19 4 17/60 0/43 28/36 5 32/68 8/56 19/56 6 49/79 9/51 16/53 7 14/26 5/23 6/9 8 22/44 10/34 9/37 9 5/18 9/18 5/19 10 2/2 1/1 0/2 TotaL 145/357 (41%) 46/274 (17%) 120/255 (47%) This table is derived from units studied in experiments with the MLF stimulating electrode placed 1.5 to 2.0 mm rostral to Deiters' nucleus and the DMLF electrode at C3. Non- projecting units include only those tested with both MLF and DMLF stimulation, and found lacking both. The sample thus tested is smaller than the samples tested for one or the other projection. (Data from ref. 3.) Vestibular Input to Cells in the Medial Nucleus Many cells in the medial nucleus can be fired monosynaptically or polysynaptically by electri- cal stimulation of the ipsilateral labyrinth (refs. 3, 4, 45, and 46). and it is this ipsilateral input that I will deal with primarily. Many cells in the medial nucleus receive inputs from the horizontal canals. Those cells excited by ipsi- lateral and inhibited by contralateral acceleration have been called type I cells; those inhibited by ipsilateral and excited by contralateral accelera- tion have been called type II cells (cf. ref. 46). Type I and type II cells are included in the sample of cells that we have studied by means of electrical stimulation. This sample undoubtedly also includes cells activated by fibers originating in parts of the labyrinth other than the horizontal canal. In our experiments with electrical stimu- lation, of 264 cells, 58 (22 percent) responded only monosynaptically, 86 (33 percent) only poly- synaptically, 34 (13 percent) fired twice âmono- synaptically and polysynaptically âwhile 86 (33 percent) were not driven at all by the stimulus. In agreement with anatomical results which show- that after destruction of the labyrinth there is a little terminal degeneration in the caudal region of the medial nucleus (refs. 16 and 19). we found that in this region (areas 1 to 4) there were very few cells driven monosynaptically by electrical stimulation of the labyrinth (fig. 5). In contrast, cells fired polysynaptically were scattered throughout the nucleus, in approximately similar proportion in all areas. It will be noticed that the distribution of cells driven monosynaptically is essentially the same as the distribution of cells with ascending or descending axons. In this region of overlap, cells of all types were fired by labyrinthine stimulation. Monosynaptic firing was observed in 32 of 40 DMLF cells, 37 of 99 MLF cells, and 27 of 106 cells lacking a long axon; polysynaptic activa- tion was observed in 9 of 35 DMLF cells. 51 of 101 MLF cells, and 54 of 94 cells without long axons. There are several interesting aspects to these results. First, it is apparent that while all types of cells receive a monosynaptic input from the labyrinth, a particularly high proportion of DMLF cells receive such an input despite the fact that MLF and DMLF cells are inter-
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAR NUCLEI 153 mingled. This gives support to the suggestion (cf. ref. 47) that while the location of a cell is of great importance in determining its input, the function of the cell also plays a role in regulating this input. Second, since many fibers in the MLF are known to end in the motor nuclei of the eye muscles, labyrinthine activation of cells projecting rostrally provides a direct reflex pathway for deviation of the eyes following move- ment of the head. Labyrinthine activation of cells projecting caudally provides a direct path- way from labyrinth to cervicothoracic cord. What type of spinal cells these caudally pro- jecting fibers end on, and what effect they have on those cells, is not known. Most investiga- tions on the role of descending vestibular fibers have concentrated on the vestibulospinal tract, and the vestibular projection in the medial longitudinal fasciculus needs further attention. Third, many cells excited by stimulation of the ipsilateral labyrinth lack long axons, and, as stated above, there are probably many inter- neurons among them. In connection with these cells, it is necessary to discuss briefly the effects of stimulation of the contralateral labyrinth. These effects, excitatory and inhibitory, were described in some detail by Shimazu and Precht (ref. 48), and many of their findings were con- firmed by ours (ref. 4). For purposes of the present discussion, it is sufficient to restate that some cells in the medial nucleus can be excited, others inhibited, by stimulation of the contra- lateral labyrinth, via commissural vestibular fibers (ref. 44); the latency of the inhibition is sometimes as short as 1.6 to 2.1 msec (ref. 4), and, as there are no crossed primary fibers (refs. 16 and 17), it seems that some of the com- missural fibers are inhibitory, and that the simplest commissural inhibitory pathway consists of an inhibitory commissural cell, activated monosynaptically by vestibular afferents (refs. 4, 49, and 50). In other cases, however, the pathway probably consists of an excitatory commissural cell that in turn excites an inhibitory neuron (probably a type II cell) located on the contralateral side, near the cell to be inhibited. It is apparent that there are inhibitory, and probably excitatory, interneurons in the medial nucleus and these cells are undoubtedly included in our sample of cells without long axons; type II cells excited by stimulation of the contralateral labyrinth usually lack long axons (refs. 4 and 51). It is reasonable to assume that the interneurons influence the activity of cells that receive ipsi- lateral excitatory vestibular inputs and that relay this input rostrally or caudally. It is therefore of interest that inputs from the ipsilateral and contralateral labyrinths often converge on cells without long axons and that, as we shall see below, the converging inputs also include some of nonvestibular origin. As is the case with Deiters' nucleus, the medial nucleus receives inputs from more than one part of the labyrinth. It seems that all three semicircular canals, as well as the utricle, supply afferent fibers to the medial nucleus (refs. 18 and 19). In experiments utilizing nat- ural stimulation, cells in the medial nucleus have so far been shown to be excited (or inhibited) by horizontal acceleration (refs. 46 and 52) and static tilting (refs. 7 and 21, and unpublished observations by B. W. Peterson). Our experi- ments do not indicate which vestibular receptors impinge on rostrally and caudally projecting cells. It is known, however, that many cells excited by horizontal acceleration have axons ascending in the MLF (ref. 53), while few have axons projecting to the spinal cord (ref. 51). So far there is little to indicate whether cells affected by tilting can be fired antidromically by spinal cord stimulation (unpublished obser- vations by B. W. Peterson). Somatic Input to Cells in the Medial Nucleus and Its Convergence With the Vestibular Input There is much less information about the somatic input to medial nucleus cells than there is about this input to Deiters' cells. A few direct spinovestibular fibers end in the caudalmost part of the nucleus (refs. 25 and 26). and from the observations of Lorente de No (ref. 24), it ap- pears that collaterals of some cerebellopetal fibers also terminate in the medial nucleus. Many of the cells studied by Frederickson, Schwarz, and Kornhuber (ref. 36) were located in the medial nucleus and, as described above in the discussion of Deiters' cells, they could be activated by deep somatic stimuli, particularly
154 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION by joint movement. All these cells also re- sponded to polarization of the labyrinth. In our experiments (ref. 4) synaptic activation from the spinal cord was often looked for only by stimulation through the same electrode (located in the cervical spinal cord near the mid- line) that was used for antidromic activation of medial nucleus cells; in some cases stimula- tion of peripheral nerves was also attempted. A number of cells were excited transsynap- tically by the cord electrode and by strong shocks (above group III threshold) to peripheral nerves, in a manner similar to that shown in figure 6. As we did not search systematically for such activation in all cells, only limited con- clusions can be drawn from the results. Most interesting is that of 38 cells in which spinal synaptic activation was seen, only 5 had long axons. Apparently somatic activation of cells in the medial nucleus is most common among interneurons. This result is similar to that of Precht and his collaborators (ref. 51), who MLF DMLF FIGURE 6.âSynaptic activation of a cell in area 3 of the medial nucleus. Extracellular recording. This cell could not be driven antidromically. Stimulation of the MLF (left) produced transsynaptic firing at a variable latency of 1.6 to 2.9 msec. Stimulation of the DMLF (right) caused firing at 1.1 to 1.7 msec. Calibrations: 500 \t.V and 1 msec. (From ref. 4.) rarely found spinal synaptic activation of type I neurons receiving an excitatory input from the ipsilateral horizontal canal, but found it fre- quently among type II cells, excited from the contralateral horizontal canal. Few of these cells could be activated antidromically from the spinal cord, and it was presumed that there were inhibitory interneurons among them. Not only somatic stimuli of spinal origin im- pinge on the group of cells we have been discuss- ing. Cells in the medial nucleus can also be activated synaptically by fibers in the MLF (fig. 6), many of which originate in the interstitial nucleus of Cajal (refs. 53 and 54). The cells activated by this descending MLF input are type II cells, excited by contralateral horizontal ac- celeration (ref. 53). Cells excited by ipsilateral horizontal acceleration are inhibited by the de- scending fibers, and it has been suggested that the inhibition is mediated by type II cells (ref. 53). In our experiments we have observed synaptic activation of 40 medial nucleus cells as a result of stimulation of the MLF. Only two of these cells had long axons. Finally, 17 cells were activated-syjiaptically from, both MLF and. spinal cord; none of these cells had long axons. It is apparent from all of these results that vestibular and nonvestibular inputs converge on many cells in the medial vestibular nucleus. Most of these cells lack long axons, and it is likely that many are interneurons whose function it is to regulate the activity of those cells that relay information from labyrinth receptors to other levels of the central nervous system. The functional role of the medial vestibular nucleus is not clear. Considerable evidence has accumulated, however, that this nucleus is in- volved in a variety of phenomena taking place during sleep, including presynaptic inhibition of different spinal pathways and of various supra- spinal structures (ref. 55). Whatever the path- ways by which these actions are carried out, further study of the regulation of activity within the medial and descending nuclei is clearly called for. A more detailed study of nonvestibular in- puts should be part of such a study.
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAH NUCLEI 155 REFERENCES 1. WILSON, V. J.; KATO, M.; THOMAS, R. C.; AND PETERSON, B. W.: Excitation of Lateral Vestibular Neurons by Peripheral Afferent Fibers. J. Neurophysiol., vol. 29,1966, pp. 508-529. 2. WILSON, V. J.; KATO, M.; PETERSON, B. W.; AND WYLIE, R. M.: A Single Unit Analysis of the Organization of Deiters' Nucleus. J. Neurophysiol., vol. 30, 1967, pp. 603-619. 3. WILSON, V. J.; WYLIE, R. M.; AND MARCO, L. A.: Organi- zation of the Medial Vestibular Nucleus. J. Neuro- physiol., vol. 31,1968, pp. 166-175. 4. WILSON, V. J.; WYLIE, R. M.; AND MARCO, L. A.: Syn- aptic Inputs to Cells in the Medial Vestibular Nucleus. J. Neurophysiol., vol. 31,1968, pp. 176-185. 5. THOMAS, R. C.; AND WILSON, V. J.: Precise Localization of Renshaw Cells With a New Marking Technique. Nature, vol. 206,1965, pp. 211-213. 6. THOMAS, R. C.; AND WILSON, V. J.: Marking Single Neurons by Staining With Intracellular Recording Microelectrodes. Science, vol. 151, 1966, pp. 1538- 1539. 7. PETERSON, B. W.: Effect of Tilting on the Activity of Neurons in the Vestibular Nuclei of the Cat. Brain Res., vol. 6,1967, pp. 606-609. 8. POMPEIANO, O.; AND BnoDAL, A.: The Origin of Vestib- ulospinal Fibres in the Cat. An Experimental-Ana- tomical Study, With Comments on the Descending Medial Longitudinal Fasciculus. Arch. Ital. Biol., vol. 95,1957, pp. 166-195. 9. NYBERG-HANSEN, R.: AND MASCITTI, T. A.: Sites and Mode of Termination of Fibers of the Vestibulospinal Tract in the Cat. An Experimental Study With Silver Impregnation Methods. J. Comp. Neurol., vol. 122, 1964, pp. 369-387. 10. PETRAS, J. M.: Cortical, Tectal and Tegmental Fiber Connections in the Spinal Cord of the Cat. Brain Res., vol. 6,1967, pp. 275-324. 11. LUND, S.; AND POMPEIANO, O.: Descending Pathways With Monosynaptic Action on Motoneurones. Ex- perientia, vol. 21,1965, pp. 602-603. 12. WILSON. V. J.; AND YOSHIDA. M.: Vestibulospinal and Reticulospinal Effects on Hindlimb, Forelimb and Neck Alpha Motoneurons. Proc. Nat. Acad. Sci., vol. 60,1968, pp. 836-840. 13. BRODAL, A.; POMPEIANO, O.; AND WALBERG, F.: The Vestibular Nuclei and Their Connections, Anatomy and Functional Correlations (Ramsay Henderson Trust Lec- tures). Oliver & Boyd, Edinburgh, London, 1962. 14. ITO, M.; HONGO, T.; YOSHIDA, M.; OKADA, Y.; AND OBATA, K : Antidromic and Transsynaptic Activation of Deiters' Neurones Induced From the Spinal Cord. Jap. J. Physiol., vol. 14,1964, pp. 638-658. 15. ECCLES, J. C.; ITO, M.; AND SZENTAGOTHAI, J.: The Cerebellum as a Neuronal Machine. Julius Springer (Berlin), 1967. 16. WALBERG, F.; BOWSHER, D.; AND BRODAL, A.: The Termination of Primary Vestibular Fibers in the Vestibular Nuclei in the Cat. An Experimental Study With Silver Methods. J. Comp. Neurol., vol. 110, 1958, pp. 391-419. 17. MUGNAINI, E.; WALBERG, F.; AND BRODAL, A.: Mode of Termination of Primary Vestibular Fibres in the Lateral Vestibular Nucleus. An Experimental Electron Micro- scopical Study in the Cat. Exptl. Brain Res., vol. 4, 1967, pp. 187-211. 18. LORENTE DE No, R.: Anatomy of the Eighth Nerve. The Central Projection of the Nerve Endings of the Internal Ear. Laryngoscope, vol. 43,1933, pp. 1â38. 19. STEIN, B. M.; AND CARPENTER. M. B.: Central Projec- tions of Portions of the Vestibular Ganglia Innervating Specific Parts of the Labyrinth in the Rhesus Monkey. Am. J. Anat., vol. 120,1967, pp. 281-318. 20. ADRIAN, E. D.: Discharge From Vestibular Receptors in the Cat. J. Physiol., vol. 101, 1943, pp. 389-407. 21. DUENSING. F.; AND SCHAEFER. K. P.: UberdieKonvergenz Verschiedener Labyrintharen Afferenzen auf Einzelne Neurone des Vestibulariskerngebietes. Arch. Psychiat. Nervenkr., vol. 199,1959, pp. 345-371. 22. FUJITA, Y.; ROSENBERG, J.; ANDSEGUNDO,J. P.: Activity of Cells in the Lateral Vestibular Nucleus as a Func- tion of Head Position. J. Physiol.. vol. 196, 1968, pp. 1-18. 23. HALLPIKE, C. S.: Some Types of Ocular Nystagmus and Their Neurological Mechanisms. Proc. Roy. Soc. Med., vol. 60,1967, pp. 1043-1055. 24. LORENTE DE No, R.: Etudes sur le Cerveau Posterieur. III. Sur les Connexions Extracerebelleuses des Fascicules Afferents au Cerveau, et sur la Function de cet Organe. Trav. Lab. Rech. Biol., vol. 22, 1924, pp. 51-65. 25. POMPEIANO, O.; AND BRODAL, A.: Spinovestibular Fibers in the Cat. J. Comp. Neurol., vol. 108, 1957, pp. 353-382. 26. BRODAL, A.; AND ANGAUT, P.: The Termination of Spino- vestibular Fibres in the Cat. Brain Res., vol. 5, 1967, pp. 494-500. 27. POMPEIANO, O.; AND Corn, E.: Analisi Microelet- trodica Delle Proiezioni Cerebellodeitersiane. Arch. Sci. Biol. Bologna, vol. 43,1959, pp. 57-101. 28. GIAQUINTO, S.; POMPEIANO, O.; AND SANTINI, M.: Riposta de Unita Deitersiane a Stimolazione Graduata di Nervi Cutanei e Muscolari in Animali Decerebrati a Cerevelleto Integro. Boll. Ital. Biol. Sper., vol. 39,1963. pp. 524-527. 29. ITO, M.; OBATA, K.; AND OCHI, R.: The Origin of Cerebellar-Induced Inhibition of Deiters' Neurones. II. Temporal Correlation Between the Transsynaptic Activation of Purkinje Cells and the Inhibition of Deiters' Neurones. Exptl. Brain Res., vol. 2, 1966, pp. 350-364. 30. ITO, M.; AND YOSHIDA, M.: The Origin of Cerebellar- Induced Inhibition of Deiters' Neurones. I. Mono- synaptic Initiation of the Inhibitory Postsynaptic Po- tentials. Exptl. Brain Res., vol. 2, 1966, pp. 330-349.
156 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION 31. POMPEIANO. O.; AND SwETT. J. E.: Actions of Graded Cutaneous and Muscular Afferent Volleys in the Decerebrate, Cerebellectomized Cat. Arch. Ital. Biol., vol. 101,1963. pp. 552-583. 32. OSCARSSON. O.: Functional Organization of the Spino- and Cuneocerebellar Tracts. Physiol. Rev., vol. 45. 1965, pp. 495-522. 33. ARMSTRONG. D. M.; ECCLES. J. C.: HARVEY, R. J.; AND MATTHEWS, P. B. C.: Responses in the Dorsal Ac- cessory Olive of the Cat to Stimulation of Hind Limb Afferents. J. Physiol.. vol. 194, 1968, pp. 125-146. 34. ECCLES. J. C.: The Way in Which the Cerebellum Proc- esses Sensory Information. Neurophysiological Basis of Normal and Abnormal Motor Activities, M. D. Yahr and D. P. Purpura. eds., Raven Press, 1967, pp. 379- 406. 35. BARKER, D.: IP. C.; AND ADAL. M. N.: A Correlation Between the Receptor Population of the Cat's Soleus Muscle and the Afferent Fibre-Diameter Spectrum of ihe Nerve Supplying It. Symposium on Muscle Re- ceptors, D. Barker, ed.. Hong Kong University Press, 1962, pp. 257-262. 36. FREDERICKSON. J. M.; SCHWARZ, D.: AND KORNHUBER, H. H.: Convergence and Interaction of Vestibular and Deep Somatic Afferents Upon Neurons in the Vestib- ular Nuclei of the Cat. Acta Oto-Laryngol., vol. 61, 1966, pp. 168-188. 37. MAGNUS. R.: Some Results of Studies in the Physiology of Posture, II. Lancet, vol. 211, 1926, pp. 585-588. 38. BATINI. C.; MORUZZI, G.; AND POMPEIANO.. O.: Cere- bellar Release Phenomena. Arch. Ital. Biol.. vol. 95, 1957, pp. 71-95. 39. BRODAL, A.; AND POMPEIANO, O.: The Origin of As- cending Fibres of the Medial Longitudinal Fasciculus From the Vestibular Nuclei. An Experimental Study in the Cat. Acta Morphol. Neerl.-Scand., vol. 1, 1957, pp. 306-328. 40. MCMASTERS. R. E.; WEISS, A. H.; AND CARPENTER, M. B.: Vestibular Projections to the Nuclei of the Ex- traorular Muscles. Degeneration Resulting From Discrete Partial Lesions of the Vestibular Nuclei in the Monkey. Am. J. Anat., vol. 118, 1966, pp. 163-194. 41. NYBERG-HANSEN, R.: Origin and Termination of Fibers From the Vestibular Nuclei Descending in the Medial Longitudinal Fasciculus. An Experimental Study With Silver Impregnation Methods in the Cat. J. Comp. Neurol., vol. 122.1964, pp. 355-367. 42. WILSON. V. J.; WYLIE. R. M.; AND MARCO, L. A.: Pro- jection to the Spinal Cord From the Medial and De- scending Vestibular Nuclei of the Cat. Nature, vol. 215. 1967, pp. 429-430. DISCUSSION Nyberg-Hansen: Concerning the medial nucleus, I noticed from one of your illustrations that there is little evi- dence of axons dichotomizing, sending one branch in the rostral and another in the caudal direction. However, we 43. LORENTEDE No, R.: Vestibulo-Ocular Reflex Arc. Arch. Neurol. Psychiat., vol. 30,1933, pp. 245-291. 44. LADPLI, R.; AND BRODAL, A.: Experimental Studies of Commissural and Reticular Formation Projections From the Vestibular Nuclei in the Cat. Brain Res., vol. 8.1968, pp. 65-96. 45. PRECHT. W.; AND SHIMAZU. H.: Functional Connections of Tonic and Kinetic Vestibular Neurons With Primary Vestibular Afferents. J. Neurophysiol., vol. 28, 1965, pp. 1014-1028. 46. SHIMAZU, H.; AND PRECHT, W.: Tonic and Kinetic Re- sponses of Cat's Vestibular Neurons to Horizontal Angular Acceleration. J. Neurophysiol., vol. 28, 1965, pp. 991-1013. 47. THOMAS, R. C.; AND WILSON. V. J.: Recurrent Inter- actions Between Motoneurons of Known Location in the Cervical Cord of the Cat. J. Neurophysiol., voL 30, 1967, pp. 661-674. 48. SHIMAZU. H.; AND PRECHT. W.: Inhibition of Central Vestibular Neurons From the Contralateral Labyrinth and its Mediating Pathway. J. Neurophysiol., vol. 29, 1966, pp. 467-492. 49. KASAHARA, M.; MANO. N.; OSHIMA, T.; OZAWA. S.: AND SHIMAZU, H.: Contralateral Short Latency Inhibition of Central Vestibular Neurons in the Horizontal Canal System. Brain Res., vol. 8, 1968, pp. 376-378. 50. MANO, M.; OSHIMA, T.; AND SHIMAZU, H.: Inhibitory Commissural Fibers Interconnecting the Bilateral Vestibular Nuclei. Brain Res., vol. 8, 1968, pp. 378-382. 51. PRECHT. W.; GRIPPO, J.; AND WAGNER. A.: Contribution of Different Types of Central Vestibular Neurons to the Vestibulospinal System. Brain Res., vol. 4, 1967, pp. 119-123. 52. DUENSING. F.; AND ScHAEFER. K. P.: Die Aktivital Einzelner Neurone im Bereich der Vestibulariskeme bei Horizontalbeschleunigungen Unter Besonderer Berucksichtigung des Vestibularen Nystagmus. Arch. Psychiat. Nervenkr., vol. 198, 1958, pp. 225-252. 53. MARKHAM, C. H.; PRECHT, W.; AND SHIMAZU. H.: Effect of Stimulation of Interstitial Nucleus of Cajal on Ves- tihular Unit Activity in the Cat. J. Neurophysiol.. vol. 29, 1966. pp. 493-507. 54. POMPEIANO. O.; AND WALBERG. F.: Descending Connec- tions to the Vestibular Nuclei. An Experimental Study in the Cat. J. Comp. Neurol., vol. 108, 1957, pp. 465-503. 55. POMPEIANO. O.: Sensory Inhibition During Motor Ac- tivity in Sleep. Neurophysiological Basis of Normal and Abnormal Motor Activities, M. D. Yahr and D. P. Purpura, eds., Raven Press, 1967, pp. 323-373. know from the work of Cajal, and from Lorente de No, too. 1 think, of such dichotomizing axons from the medial nucleus. Do you have any comments on this? Wilson: My first comment is that we did find far fewer
INPUTS TO CELLS OF THE LATERAL AND MEDIAL VESTIBULAR NUCLEI 157 dichotomizing fibers than we expected. This may be ac- counted for to some extent by our missing some of the rostrally and caudally descending fibers. The other thing you must realize is that I am speaking about long dichotomiz- ing axons which can be stimulated in the midline and in the spinal cord. Some of the axons may go up in some other direction where we fail to stimulate one of the branches. Also from my experience, looking back at Cajal's and Lorente's pictures, it seems that it is sometimes a little hard to know exactly where the cell bodies are. I am not com- pletely sure what the main reason is for our finding relatively few dichotomous fibers. Perhaps there are just fewer of these than we expected. Nyberg-Hansen: How far rostrally did you stimulate the medial longitudinal fasciculus? Wilson: We put our electrode approximately 2 mm rostral to Deiters' nucleus. I think this would be rostral to the end of the medial nucleus. Borison: Do you have any reason to exclude the forebrain from participation? That is to say, have the latencies of effects and the time courses given you reason to exclude pathways from distant sources reaching, say, the cerebral cortex? In this connection do you think your results might have been different in unanesthetized decerebrate animals? Finally, did you curarize your animals, and have you any reason to believe that there might not have been secondary effects due to reflex responses in the periphery? Wilson: As far as curarization is concerned, of course the animals were Flaxedilized, and there has been no move- ment. Secondly, it is likely that, if I had done my experi- ments on animals that were not anesthetized and decerebrate, or under Nembutal. or under any one of many other possible conditions, some results would have been different. Con- cerning the results of the monosynaptic inputs from the labyrinth, the location of cells driven antidromically, and anything else of this type, there would be no difference what- soever. One thing which you can consider is these long- latency and diffuse effects coming up from the spinal cord. They were similar in animals under Nembutal and under chloralose anesthesia. I suspect they would be reasonably similar under other conditions, but not exactly the same. Probably the patterns of activity would have been different under some conditions. As far as exclusion of effects going rostrally and then coming back is concerned, the latencies of the facilitation, for example from the forelimbs to Deiters' nucleus, are on the order of a few milliseconds. I really do not believe there is enough time for this to go all the way to the cortex and come back again, but certainly I cannot exclude that there is something going on rostrally to the vestibular nuclei. I do not think it is a major factor. Ito: Relevant to your observation on the influence of group la muscle afferents upon Deiters' neurons, have you tested the effect of joint afferents which, in the previous work by Gemandt, Livingston, and Katsuki. seem to have a powerful action on Deiters' neurons? Wilson: No. So far we have not specifically dissected out any joint nerves to stimulate them. Of course it is possi- ble that some of the nerves that we stimulated there were contaminated by joint afferents. This is something I cannot really discuss, but it is part of the work we are doing now. By the way, I still find it hard to believe that there is no la input, direct or indirect, to Deiters' nucleus. Pompeiano: 1 should like to discuss briefly two problems which are relevant to the nice presentation made by Dr. Wilson. The first problem concerns the effects of stimula- tion of the primary afferents on Deiters' neurons. The sec- ond problem concerns the effects of stimulation of the ves- tibular nerve on the primary afferents in the spinal cord. With respect to the peripheral influences on Deiters' neurons, our observations made in unanesthetized decerebrate cats with the cerebellum intact clearly support the conclusion of Dr. Wilson. In particular, we were unable to find any change in the activity of the Deiters' neurons on repetitive stimula- tion of the group la afferents (Giaquinto, S.: Pompeiano. O.: and Santini, M.: Response of Deitersian Units to Graduated Stimulation of Cutaneous and Muscular Nerves in Decere- brate Animals with Intact Cerebellum. Boll. Soc. Ital. Biol. Sper., vol. 39, 1963, pp. 524-527). On the other hand, the cutaneous and the high-threshold muscular afferents exerted a clear-cut effect on Deiters' neurons. On the basis of these findings, it may be questioned whether Deiters' neurons receive collaterals from the dorsal spinocerebellar tract (DSCT), at least from that subdivision of the DSCT which transmits group la volleys to the cerebellum. In our experi- ments the response patterns of Deiters' neurons to stimula- tion of the cutaneous and high-threshold muscular afferents were generally characterized by a facilitation and in a few in- stances by inhibition; however, with an increase in stimulus intensity, the majority of the units responded in a complex fashion. The most common response was an initial short- latency discharge, followed by a silent period or a late dis- charge. It seems likely, at least in the preparation with the cerebellum intact, that some competition of facilitatory and inhibitory effects occurs at the level of the vestibular neurons from converging afferent volleys of different spatial and temporal dispersions. With respect to the second problem, I should like to men- tion that stimulation of the vestibular nerve performed in decerebrate cats evokes dorsal-root potentials in the lumbar cord at the time that descending vestibular volleys elicit motoneuronal discharges (Cook, W. A., Jr.; Cangiano, A.; and Pompeiano, O.: Vestibular Influences on Primary Ef- ferents in the Spinal Cord. Pfliigers Arch, ges., Physiol., vol. 299, 1968, pp. 334-338). The vestibular evoked primary afferent depolarization involved group I afferents from both extensor and flexor muscles and also large group II cutaneous afferents. These findings suggest that when spinal motoneu- rons are triggered by the vestibular apparatus, it may be functionally important to reduce the segmental afferent input to these motoneurons by the mechanism of presynaptic inhibi- tion. This partial deafferentation may prevent instabilities in the motor system, which might occur when somatic sen- sory volleys elicited during movements are fed back into the spinal cord and interact with the discharging motoneurons. Wilson: This, of course, is something I covered in mention- ing the various interesting observations made by you. Precht: Just a short comment on the function of the medial nucleus with respect to the spinal cord. We tried to activate
158 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION antidromically the vestibular neurons that responded to hori- zontal rotation, having in mind the idea that there is an effect of horizontal rotation on the spinal cord reflexes as Gernandt and Thulin showed many years ago. However, it is quite striking that almost none of the second-order neurons of the horizontal semicircular canal is activated antidromically by stimulating the spinal cord at Cs. Wilson: Type I? Precht: Type I is not activated; however, several of the type II and type III neurons were antidromically excited. The effect of horizontal rotation on spinal reflexes cannot be satisfactorily explained on the basis of these data. Prob- ably vestibuloreticulospinal projections are also of great im- portance. So we are still in the functional "no-man's land" at this particular point. We know to some extent the syn- aptology of the vestibulospinal system, but we do not know what kind of sensory information is transmitted to the spinal cord via the vestibulospinal tracts. Wilson: We have also been interested in this. Peterson has licc-n looking to see whether any of the cells that he sees in the medial nucleus are influenced by tilting. Unfor- tunately, so far he has not had good tilting effects and good .minimum driving in the same cat, but we are pursuing this also. Precht: A short comment to the problem of monosynaptic and polysynaptic activation of the vestibular neurons in re- sponse to vestibular nerve stimulation. In talking about second-order neurons of the horizontal canal, it is quite sur- prising that, if one uses the adequate stimulus, the thresh- old of the monosynaptically activated neurons is much higher, significantly higher, than that of the ones that are polysynaptically activated. So it is quite important from the standpoint of a correlation between anatomy and physiology that a monosynaptic input is not necessarily powerful in terms of assuring a high sensitivity to a given sensory input. Wilson: Absolutely. I should like to add right here that I think the same thing applies to the output side. Very often when we talk about the output, and the vestibulospinal tract is an example, we concentrate on the monosynaptic effect. It is my personal opinion that most of the effects that you see on the extensor tone are mediated polysynaptically. The monosynaptic action is a more discretely organized arrange- ment.
SESSION VI Chairman: CESAR FERNANDEZ University of Chicago