The Journal of Neurophysiology Vol. 80 No. 5 November 1998, pp. 2352-2367
Copyright ©1998 by the American Physiological Society

Neuronal Activity in the Vestibular Nuclei After Contralateral or Bilateral Labyrinthectomy in the Alert Guinea Pig

Laurence Ris and Emile Godaux

Laboratory of Neurosciences, University of Mons-Hainaut, B-7000 Mons, Belgium

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Ris, Laurence and Emile Godaux. Neuronal activity in the vestibular nuclei after contralateral or bilateral labyrinthectomy in the alert guinea pig. J. Neurophysiol. 80: 2352-2367, 1998. In the guinea pig, a unilateral labyrinthectomy is followed by an initial depression and a subsequent restoration of the spontaneous activity in the neurons of the ipsilateral vestibular nuclei. In two previous works, we have established the time course of these changes in the alert guinea pig using electrical stimulation as a search stimulus to select the analyzed neurons. The latter criterion was important to capture the many ipsilateral neurons that are silent at rest during the immediate postlabyrinthectomy stage. Because it is known that a pathway originating from the vestibular nuclei on one side crosses the midline and functionally inhibits the activity of the vestibular nuclei on the other side, we investigated in the first part of this study the spiking behavior of the neurons in the vestibular nuclei contralateral to the labyrinthectomy using the same procedure as that used for the ipsilateral neurons. The spiking behavior of 976 neurons was studied during 4-h recording sessions in intact animals and 1 h, 1 day, 2 days, or 1 wk postlabyrinthectomy. Neurons selected according to the electrical activation criterion were classified further as type I (their firing rate increased during ipsilateral rotation), type II (their firing rate increased during contralateral rotation), or unresponsive. The resting activity of type I neurons, which was 38.1 ± 20.9 spikes/s (mean ± SD) in the control state, increased statistically significantly 1 h after the lesion (53.3 ± 29.1 spikes/s) and remained at this level 1 wk later (56.0 ± 20.3 spikes/s). The sensitivity of type I units, which was 0.80 ± 0.46 spikes/s per deg/s in the control population, decreased to 0.49 ± 0.26 spikes/s per deg/s 1 h after the lesion and remained at this level 1 wk later (0.50 ± 0.39 spikes/s per deg/s). When all monosynaptically activated neurons (type I, type II, unresponsive) were pooled, the sensitivity to horizontal rotation fell from 0.58 ± 0.51 spikes/s per deg/s in the control state to 0.15 ± 0.25 spikes/s per deg/s 1 h after the lesion and to 0.20 ± 0.32 spikes/s per deg/s 1 wk later. The major findings of the first part of this study in the alert guinea pig are thus in accord with those of Curthoys et al. and Smith and Curthoys in anesthetized guinea pigs. In the second part of this work, we studied the spiking behavior of the neurons in the vestibular nuclei after bilateral labyrinthectomy. After unilateral labyrinthectomy, the resting discharge of the ipsilateral monosynaptically activated vestibular neurons fell from 36.9 ± 21 spikes/s (basal activity) to 6.7 ± 17.0 spikes/s 1 h after the lesion and then recovered, reaching 17.4 ± 18.9 and 40.8 ± 23.7 spikes/s 1 day and 1 wk after the lesion, respectively. These observations raise the two following questions. What are the relative contributions of the loss of the excitatory influence from the ipsilateral labyrinth (destroyed) and of the persistence of the inhibitory influence from the contralateral labyrinth (intact) in the labyrinthectomy-induced depression of activity? And are the left-right asymmetries caused by a unilateral labyrinthectomy the driving force for restoration of activity? Here, we addressed these two questions by studying the spiking behavior of 473 second-order vestibular neurons in the alert guinea pig after a bilateral labyrinthectomy. In the acute stage, 1 h after bilateral labyrinthectomy, the resting discharge of the second-order vestibular neurons was 16.2 ± 22.4 spikes/s. From comparison with the results obtained in the acute stage after a unilateral labyrinthectomy, we inferred that the ipsilateral excitatory influence was between two and three times more powerful than the contralateral inhibitory influence. After bilateral labyrinthectomy as well as after unilateral labyrinthectomy, the resting activity of the second-order vestibular neurons returned to normal, reaching 20.8 ± 23.1 spikes/s 1 day after the lesion and 38.6 ± 21.1 spikes/s 1 wk after the lesion. From this fact, we concluded that the left-right asymmetries caused by a unilateral labyrinthectomy were not the error signals inducing the restoration of activity.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

In mammals, unilateral labyrinthectomy induces an immediate and severe disabling in the control of the resting position of the body, the head, and the gaze. These static symptoms include a lateral deviation of the head and body toward the side of injury and a spontaneous ocular nystagmus with its slow phases directed toward the same side (Azzena 1969; de Waele et al. 1989; Fetter and Zee 1988; Jensen 1979; Lacour and Xerri 1981; Llinas and Walton 1979; Precht et al. 1966; Schaefer and Meyer 1974). The motor control of the unilaterally labyrinthectomized mammals also shows severe deficits in the motor responses to head movements such as the vestibuloocular reflexes (VOR) and the vestibulospinal reflexes (dynamic symptoms) (Baarsma and Collewijn 1975; Dutia 1985; Fetter and Zee 1988; Maioli and Precht 1985; Vibert et al. 1993; Zennou-Azogui et al. 1993). Over time, this vestibular syndrome abates in a recovery process known as vestibular compensation (for a review, see Dieringer 1995; Smith and Curthoys 1989). However, although the static symptoms of labyrinthectomy largely disappear rapidly (de Waele et al. 1989; Ris et al. 1997; Schaefer and Meyer 1974; Smith et al. 1986), the dynamic symptoms diminish much more slowly (Baarsma and Collewijn 1975; Fetter and Zee 1988; Halmagyi et al. 1990; Maioli et al. 1983; Vibert et al. 1993).

At the neuronal level, it has been reported repeatedly that the resting discharge in the neurons of the ipsilateral vestibular nuclei is reduced just after a unilateral labyrinthectomy and then recovers spontaneously (Newlands and Perachio 1990a; Pompeiano et al. 1984; Precht et al. 1966; Ried et al. 1984; Smith and Curthoys 1988b; Xerri et al. 1983). In the guinea pig, the resting discharge remains seriously depressed during 10 h, whereas its restoration is complete 1 wk after the lesion (Ris et al. 1995, 1997). The latter phenomenon plays an obvious role in the recovery from behavioral symptoms, although it is not the only factor playing a role in this process (Ris et al. 1997).

Since the experiments of Shimazu and Precht (1966), it has been known that a pathway originating from the vestibular nuclei on one side crosses the midline and functionally inhibits the activity of the vestibular nuclei on the other side. Hence, it could be expected that a labyrinthectomy, by causing a loss of this inhibitory influence on the contralateral vestibular nuclei, would consequently induce an increase of activity in these nuclei. However, there are discrepancies in the literature about what occurs in the contralateral (intact side) vestibular nuclei. The most-studied neurons have been the type I neurons (increasing their firing rate during rotation toward the recording side) of the medial vestibular nucleus (MVN). After a unilateral labyrinthectomy, the resting activity of those contralateral neurons was shown to be increased in the cat (Markham et al. 1977) and in the guinea pig (Curthoys et al. 1987, 1988; Smith and Curthoys 1988a), whereas other studies reported that there was no such change either in the albino rat (Hamann and Lannou 1988) or in the gerbil (Newlands and Perachio 1990a). Even worse, in total contradiction with these results, there have been reports that in the cat, 1 day after unilabyrinthectomy, both MVN were electrically silent (McCabe and Ryu 1969; McCabe et al. 1972), the shutdown of activity in the contralateral vestibular nuclei being attributed to the cerebellum because it did not occur when the animal was cerebellectomized (McCabe et al. 1972).

The first aim of this study was to investigate in the alert guinea pig with an intact CNS the spontaneous activity in the contralateral vestibular nuclei at different times after a unilateral labyrinthectomy. Furthermore in an attempt to better understand the mechanisms of the dynamic vestibular symptoms and of their slow recovery, we analyzed the responses to horizontal rotation in all the studied neurons.

From the existence of functionally inhibitory commissural connections (Shimazu and Precht 1966), a second prediction can be made: the low resting discharges observed in the ipsilateral vestibular nuclei after a unilateral labyrinthectomy are due not only to the loss of the excitatory influence from the ipsilateral labyrinth but also to the persistence of the inhibitory influence from the contralateral labyrinth. The second aim of this paper was to separate the two components of this phenomenon by comparing the resting activity of the neurons in the vestibular nuclei just after a labyrinthectomy performed either bilaterally (the present work) or unilaterally on the ipsilateral side (data collected in previous works, Ris et al. 1995, 1997). The result will be important to interpret data concerning resting activity of vestibular neurons collected in vitro in brain slices or in the isolated brain stem. Indeed, in these two approaches, a bilateral labyrinthectomy is, of course, performed when brain stem is removed.

On the other hand, the spontaneous restoration of a resting activity in the vestibular neurons deprived of their labyrinthine input raises the following important question. What is the error signal that stimulates that restoration of activity? It could be that the silenced neurons themselves would detect either their own absence of spiking activity or the loss of activity in the primary vestibular fibers connected to them. Another hypothesis is that the driving force for restoration of activity would be the left-right asymmetry either among the activities of the vestibular nuclei or among the motor commands inducing the symptoms or among the afferent signals resulting from the symptoms. The third aim of this work was to test the latter hypothesis by studying whether restoration of activity still could occur in vestibular neurons after a bilateral labyrinthectomy.

	METHODS

Abstract Introduction Methods Results Discussion References

Experimental data were obtained from 24 pigmented female guinea pigs weighing between 400 and 800 g obtained from an authorized supplier (Charles River, France). All experiments conformed to the recommendations of the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health. All operations were performed under halothane anesthesia and aseptic conditions.

General design of the experiments

During a first (preliminary) operation, all guinea pigs were prepared for chronic recording of extracellular spikes of the neurons in the rostral half of the left vestibular nuclear complex and for chronic recording of eye movements. One or 2 wk after this first operation the experiments began. The animals were divided into four groups of six individuals. In the first group (control), the labyrinths were left intact and single-neuron recordings were made during one 4-h session. In the second group, single-neuron recordings began 1 h after a right labyrinthectomy and were made during two 4-h sessions separated by a rest period of 1 h. In the third group, recordings were made on days 1, 2, and 7 after the contralateral lesion. During each of these days, neurons were recorded during a single 4-h session. In the fourth group, single-neuron recordings were made after a bilateral labyrinthectomy during three 4-h sessions: the first starting 1 h after the lesion, the second 1 day after the lesion, and the third 1 wk after the lesion.

In addition, control data collected here were pooled with those obtained in six control animals in a previous work (Ris et al. 1995). The spiking behaviors of the units (which were recorded in the same way as in the present experiments) from the latter study were determined again using the same analysis procedure as that used in the present work (see Data analysis).

Preliminary operation

In the preliminary operation, guinea pigs were fitted with several chronic devices. A 9-mm-diam coil made of three turns of Teflon-coated seven-stranded stainless steel wire was implanted subconjunctivally on the right eye (Judge et al. 1980). A head holder was cemented to the skull. A craniotomy was performed over the cerebellum (4 mm wide and 5 mm long; stereotaxic coordinates: L 0-4 mm, P 5-10 mm) (Rapisarda and Bacchelli 1977). The dura mater was removed, and a dental cement chamber constructed around the hole. Between the recording sessions, the surface of the cerebellum was protected with a silastic sheet and the chamber sealed with bone wax. To enable stimulation of the left vestibular nerve in control animals and in guinea pigs, which would undergo contralateral labyrinthectomy later, two electrodes consisting of two Teflon-coated silver wires, denuded at their tips (ball electrodes) were placed over the round window of the middle ear cavity and in front of the horizontal and anterior semicircular ampullae, respectively. The leads from the coil and from the stimulating electrodes were passed subcutaneously toward the head region where they were soldered onto sockets cemented to the holding system.

Labyrinthectomy

The procedure to perform a global labyrinthectomy was as follows. Lateral and anterior semicircular canals were approached via the dorsal bulla, whereas the posterior semicircular canal and the cavity containing the utricle and saccule were reached via the mastoid bulla. The different parts of the membranous labyrinth then were extirpated one after the other. At the end of bilateral labyrinthectomy, two electrodes were placed on the left side, one over the round window and the other one on the site of the destroyed labyrinth, to enable stimulation of the left vestibular nerve. For this operation, halothane had to be administered for only 45 min. After such short anesthesia, the animals recovered a normal state of alertness ~20 min after cessation of halothane delivery. During the first hour after awakening, the animal exhibited severe imbalance. To minimize discomfort and to prevent the animal from hurting itself, it was allowed to recover in the corner of a small box (30 × 30 × 15 cm) the floor and walls of which were covered with padding.

Eye-movement recording

Eye movements were measured using the scleral search coil technique (Fuchs and Robinson 1966). The measurement system had a bandwidth of 1,000 Hz and a sensitivity of 0.25°. Calibration was obtained by rotating both magnetic fields ±5° around the horizontal and vertical axes with the guinea pig's head kept still in space. Transitivity was checked by substituting a coil fitted on a pivot for the guinea pig eye. Just before each 4-h session of neuronal recording, the zero position of the gaze had to be estimated because some vestibular neurons were sensitive to eye position. For this purpose, the vertical and horizontal positions of the right eye were sampled at a rate of 10/s during spontaneous ocular movements made in the light during a period of 10 min. Zero position was obtained by computing the mean horizontal and vertical positions of gaze (n = 6,000 ocular positions).

Recording of neuronal activities

Each experimental session began by attaching the animal's head to a holding bar located in the center of a turntable and placed so that the plane defined by the two horizontal semicircular canals was coincident with the earth horizontal plane (head pitched 40° nose down) (Curthoys et al. 1975). Special caution was taken to avoid any discomfort in the animals throughout the experimental sessions. In particular, the body of the animal was wrapped in a light cover and immobilized using small cushions. The effect of this procedure on the animal was assessed by monitoring the heart rate and respiratory rhythm. The heart rate was recorded with electrodes positioned on the skin of the thorax. Respiratory rhythm was recorded by means of a thermal probe located near a nostril, which detected changes in air temperature as air was expired and inspired. Our immobilization procedure did not affect heart rate or respiratory rate. Moreover, shortly after labyrinthectomy, holding the head of the animal alleviated stress associated with imbalance. The cranial opening was cleaned with sterile saline and antibiotics. Furthermore, local anesthetics were used to irrigate the cement chamber to prevent any pain. Before the start of the recording session of neuronal activity, the vestibular nuclear complex was localized electrophysiologically (Fig. 1, A and B). A glass microelectrode (1-5 M impedance at 1,000 Hz), attached to a micromanipulator, was vertically lowered through the cranial opening in the direction of the left vestibular nuclei. The N1 and N2 waves of the field potential evoked by stimulation of the vestibular nerve (de Waele et al. 1990; Newlands and Perachio 1990a; Ris et al. 1995; Shimazu and Precht 1965) were used to map out the location of the rostral part of the vestibular complex.

View larger version (23K): [in this window] [in a new window]   FIG. 1. General procedure used for recording the resting activity of the neurons in the vestibular nuclei of the alert guinea pig after a contralateral or a bilateral labyrinthectomy. A: sketch of the preparation. A cranial opening allowed a micropipette to be lowered in the direction of the left vestibular nuclei. Two chronic electrodes enabled stimulation of the left vestibular nerve. A right labyrinthectomy or a bilateral labyrinthectomy was performed 1 h to 1 wk before recording. B: field potential (consisting of 2 waves, N1 and N2) evoked within the vestibular nuclear complex by an electric shock applied on the ipsilateral vestibular nerve. This field potential was used to localize the vestibular nuclei. C: illustration of the activation of a neuron by an electric shock applied on the vestibular nerve. The responses to four successive stimulations are superimposed. The "all-or-none" characteristic of the response demonstrated that it corresponded to a single unit and not to a field potential.

Because we wanted to compare the behavior of neuronal populations recorded in different groups of animals and at different postoperative times in the same animals, the studied neuronal assemblies had to be as similar as possible. We therefore defined criteria to select the neurons included in this work. 1) The analyzed neurons had to be recorded in the anterior part of the vestibular complex, and 2) they had to be recruited by an electric stimulation of the ipsilateral nerve. To record neuronal activity in the vestibular nuclei, a glass micropipette was lowered in the region where the vestibular field potential was localized previously. The left vestibular nuclei then were explored during stimulation of the left vestibular nerve. The intensity of the square pulses used as a search stimulus was between two and three times the threshold intensity required to obtain an N1 potential. The small amplitude of the field potential obtained with such pulses did not obscure evoked single action potentials. The "all-or-none" characteristic of the response demonstrated that it corresponded to a single unit and not to a field potential (Fig. 1C). Once a unit was recruited, its threshold was determined to be the shock level where the neuron followed ~50% of the pulses. The latency of the response then was measured with a stimulation of twice the unit threshold. The activity of the neurons so identified then was recorded in complete darkness while keeping the animal still in space and while rotating it horizontally and sinusoidally at 0.3 Hz with a peak angular velocity of 40°/s for 20 cycles. The alertness of the animals was maintained by producing unexpected sounds.

Data analysis: general procedure

Analysis was performed off-line on PC 486 IBM-compatible computers after storing on disk data from digital audio tape recordings. For each studied neuron, we determined its resting activity and its sensitivity to head velocity during rotation in the horizontal plane.

The resting activity was obtained by measuring the mean firing rate of the spikes occurring during 1 min (true spontaneous activity).

The sensitivity to horizontal head rotation was determined according to the following procedure. The horizontal and vertical components of the eye position and the turntable-velocity signal were sampled at 100 Hz. The VOR induced by a sinusoidal rotation of the turntable consists of slow phases separated by quick resetting phases (Fig. 2A). Identification of the slow phases was performed manually using an interactive algorithm. Unitary activity of the recorded neurons was amplified and transferred to a window discriminator. The time axis of this signal was divided into intervals 250 µs wide. During each interval, the presence or absence of an action potential was checked. For each occurrence of an action potential, the instantaneous firing rate was calculated as the inverse of the interspike interval immediately preceding that action potential. Finally, the instantaneous firing rate corresponding to the sampled eye and head signals (at every 10 ms) was obtained by interpolation. A multiple linear regression then was performed with instantaneous firing rate (z) as the dependent variable and horizontal head velocity (x) and horizontal eye position (y) as the independent variables (Fig. 2B). The slope of the intercept line between the regression plane and the x-z plane (firing rate-head velocity plane) gives the sensitivity of the neuron to head velocity. Eye position was not used as an independent variable when its correlation with firing rate was not statistically significant (P > 0.001). In the latter case, a linear regression was performed with instantaneous firing rate as the dependent variable and head velocity as the independent variable. The slope of this regression line then was interpreted as the sensitivity of the neuron to head velocity. Null values of the firing rate were not taken into account. Thanks to this precaution, the method remained valid when the rotation-induced firing rate modulation was asymmetric with a cutoff of the discharge when the head was rotated in one direction. When the neuron changed its discharge in relation to quick phases, interference from this behavior was avoided by disregarding data collected during the quick phases, during the 100 ms preceding each of them, and during the 100 ms after each of them.

View larger version (19K): [in this window] [in a new window]   FIG. 2. Illustration of the method used to calculate the sensitivity to horizontal head velocity of each studied vestibular neuron. A: example of simultaneous measurements of horizontal eye position, head velocity, and firing rate. B: firing rate is plotted as a variable (z axis) dependent on 2 independent variables: head velocity (x axis) and eye position (y axis). Regression plane is calculated by multiple linear regression. Its intercept line with the x-z plane is determined. Slope of the latter line, r, gives the sensitivity of the neuron to head velocity (spikes/s per deg/s).

The methodology described here above was slightly different from that used in our previous works on the activity of the vestibular nuclei after an ipsilateral labyrinthectomy (Ris et al. 1995, 1997). Because it is simpler and more widespread, we decided to use it here and in future studies. However, in order not to invalidate comparison of the present results with the previous ones, the parameters of the spiking behavior of each neuron recorded in the previous studies were redetermined by the procedures used in the present study. Those values are given in Table 2 (see DISCUSSION).

Statistical analysis

In this study, we needed to compare statistically the parameters of populations of neurons recorded in different groups of animals. As the two main parameters (resting discharge and sensitivity to rotation) were not normally distributed, we used nonparametric tests. The Wilcoxon rank sum test was used to compare two populations of neurons. The Kruskal-Wallis test was used to estimate the evolution of a parameter as a function of the time elapsed after uni- or bilateral labyrinthectomy by comparing more than two populations.

Histology

After completion of recording sessions, three small electrolytic lesions 1 mm apart (30 µA for 10 min) were made under the vestibular nuclei using a glass microelectrode. Then the animals were anesthetized deeply with pentobarbital sodium and transcardially perfused with 0.9% saline followed by 10% formaldehyde. The locations of the electrolytic lesions were verified on 20 µm transversal sections tilted 40° posterior with respect to the vertical and stained with cresyl fast violet (see Fig. 3 in Ris et al. 1995). The anatomic location of each brain stem unit was established by combining the histological controls and micrometer readings with the aid of the map of the vestibular nuclear complex established in the guinea pig by Gstoettner and Burian (1987) and slightly modified here (see DISCUSSION).

View larger version (59K): [in this window] [in a new window]   FIG. 3. Location of the neurons recorded in control animals and after contralateral labyrinthectomy. A-E: subdivisions of the rostral part of the vestibular nuclear complex. Plane of the illustrated sections is tilted 40° posteriorly with respect to the vertical stereotaxic plane as defined by Rapisarda and Bacchelli (1977). Reference point is the center of the abducens nucleus, which has a diameter of only 0.25 mm in the guinea pig. A and B: sections are 600 and 360 µm anterior to the center of the abducens nucleus, respectively. C: section passes through the reference point. D and E: sections are 440 and 800 µm posterior to the reference point. S, superior vestibular nucleus; M, medial vestibular nucleus; Lv, ventral part of the lateral vestibular nucleus; Ld, dorsal part of the lateral vestibular nucleus; D, descending vestibular nucleus; 8n, vestibular nerve; I, interstitial nucleus of the vestibular nerve; X, X group; Y, Y group; 4th v, fourth ventricle; 7n, facial nerve; 7g, genu of the facial nerve; bc, brachium conjunctivalis; C, cochlear nucleus; 5d, descending root of the trigeminal nucleus; rb, restiform body; sa, stria acustica; PH, nucleus prepositus hypoglossi; 6, abducens nucleus. F-K: distribution of the recorded neurons in the S, M, Lv, Ld, and D subdivisions during the different recording periods. In each panel, G-K, the time indicated (top right) corresponds to the starting time of the period after the labyrinthectomy during which recordings were made.

	RESULTS

Abstract Introduction Methods Results Discussion References

For the sake of clarity, results will be presented in two separate sections. All data collected in animals having undergone a contralateral (right) labyrinthectomy will be described first, whereas all observations made in bilaterally labyrinthectomized animals will be reported afterward.

Neuronal activity after contralateral labyrinthectomy

GENERAL CHARACTERISTICS OF THE STUDIED NEURONS. In this study, as in our previous ones (Ris et al. 1995, 1997), we studied a population of vestibular neurons characterized by their location in the anterior half of the left vestibular nuclear complex and by their activation by an electric stimulation of the ipsilateral vestibular nerve.

Recordings were made from 817 electrically recruited neurons in intact guinea pigs and in animals that had undergone a right labyrinthectomy from 1 h to 1 wk beforehand. Each recording session lasted 4 h, and the spiking parameters of the neurons recorded in the same 4-h session were combined. To avoid lesions due to repeated penetrations in the same vestibular nuclear complex, the control recording session, the early postlabyrinthectomy recording sessions (beginning 1 and 6 h after the lesion), and the later recording sessions (24 h, 48 h, and 1 wk after the lesion) were performed on three distinct groups of six individuals. In addition, the data concerning the activity of 159 electrically recruited neurons recorded in six control guinea pigs in a previous study (Ris et al. 1995) have been pooled with the control data obtained in the present study after it was checked that the spiking parameters (resting rate, sensitivity to head rotation) of both control populations were not statistically different. This had led the number of spiking behaviors analyzed in this section of this study to a total of 976.

In the anterior part of the vestibular nuclear complex, each of our electrically recruited neurons was found to be located in either the superior vestibular nucleus (S), the rostral pole of the medial vestibular nucleus (M), the ventral part of the lateral vestibular nucleus (Lv), the dorsal part of the lateral vestibular nucleus (Ld), or the rostral pole of the descending vestibular nucleus (D) (see the map presented in Fig. 3, A-E). The relative densities of recordings from each area are shown, for each recording session, in Fig. 3, F-K. These distributions show that neurons were recorded from each of the explored vestibular nuclei.

The basic criterion for a neuron to be included in this study was its activation by an electric shock applied on the vestibular nerve. They were further classified on the basis of their mono- or polysynaptic activation. The criterion used to establish monosynaptic recruitment was based on the following considerations. In the guinea pig, the peak latency of the N1 wave of the field potential evoked in the vestibular nuclei by a stimulation of the ipsilateral vestibular nerve has been consistently found to be 1 ms. On the other hand, the shortest peak latency of the unitary potentials recorded by our group was 0.85 ms. This value was 0.15 ms lower than the latency of the N1 wave peak, which was assumed to correspond to the mean monosynaptic latent period (Precht and Shimazu 1965). Hence, we considered neurons to be unequivocally monosynaptically recruited when they were activated at latencies between 0.85 and 1.15 ms. This criterion was used throughout the study. However, a unit that did not fire in the currently defined monosynaptic range could not be considered unequivocally as not receiving monosynaptic inputs. This was especially true for latent periods between 1.15 and 1.30 ms. In our sample, 79% of the units had an orthodromic activation latency between 0.85 and 1.15 ms, whereas the activation latencies of 9% of our neurons ranged between 1.15 and 1.30 ms.

In this study, responsiveness to horizontal head rotation was not used as a selection criterion but as a classification criterion. Each electrically recruited neuron was classified as an unresponsive neuron, a type I neuron (its firing rate increased during ipsilateral rotation), or a type II neuron (its firing rate increased during contralateral rotation) (Duensing and Schaefer 1958).

Table 1 classifies the studied neurons according to the period when they were recorded, their mono- or polysynaptic recruitment, their type (type I, type II, unresponsive), and their location.

BEHAVIOR OF CONTROL MONOSYNAPTICALLY RECRUITED NEURONS. Monosynaptically recruited vestibular neurons (261) were recorded in control guinea pigs. When submitted to horizontal head rotation, 123 (47%) behaved as type I, 65 (25%) as type II, whereas 73 (28%) were unresponsive.

Figure 4A shows the distribution of these neurons. When the neurons from the three types (unresponsive, type I, type II) were pooled, the resting discharge had a mean of 36.9 ± 21.0 (SD) spikes/s and ranged from 1.6 to 119.7 spikes/s. The mean resting discharge was 38.1 ± 20.9 spikes/s for type I units, 33.6 ± 20.8 spikes/s for type II units, and 37.7 ± 21.5 spikes/s for unresponsive units.

View larger version (34K): [in this window] [in a new window]   FIG. 4. Properties of the monosynaptically recruited vestibular neurons of the control group (n = 261). A: histogram of the resting discharges of the neurons. Mean is 36.9 ± 21.0 (SD) spikes/s. There is no silent neuron. Neurons unresponsive to head rotation (), neurons behaving as type I (), and neurons behaving as type II () are considered separately. B: histogram of the sensitivities of the neurons to horizontal head rotation. Mean is 0.58 ± 0.51 (SD) spikes/s per deg/s.

Figure 4B shows the distribution of the sensitivities to horizontal head rotation of this control neuronal population. When unresponsive units were pooled with type I and type II neurons, the sensitivity to head velocity had a mean of 0.58 ± 0.51 (SD) spikes/s per deg/s and ranged from 0.00 to 1.94 spikes/s per deg/s. The mean sensitivity to head velocity of type I neurons was 0.80 ± 0.46 spikes/s per deg/s, whereas that of type II neurons was 0.77 ± 0.38 spikes/s per deg/s.

BEHAVIOR OF THE MONOSYNAPTICALLY RECRUITED NEURONS JUST AFTER CONTRALATERAL LABYRINTHECTOMY. Just after the labyrinthectomy, from 1 to 5 h after the lesion, 105 monosynaptically recruited neurons were recorded. The main observations were the following: there was no shutdown of the resting discharges, there was a statistically significant increase of the mean resting rate of the type I neurons, and the sensitivity of the neurons to head rotation was seriously reduced, a fact that probably explains the larger number of encountered unresponsive neurons.

There were 70 (66%) unresponsive units, 22 (21%) type I neurons, and 13 (13%) type II neurons.

Figure 5A presents the distribution of the resting rates of the neurons. When the three types of neurons (unresponsive, type I, type II) were pooled, the mean resting was 41.2 ± 26.9 spikes/s, which was similar to that observed in the control population (36.9 ± 21.0 spikes/s; Wilcoxon test, P = 0.19; see also Fig. 6A). The mean resting rate of the type I units was larger just after labyrinthectomy (53.3 ± 29.1 spikes/s) than in the control group (38.1 ± 20.9 spikes/s; Wilcoxon test, P < 0.05; Fig. 6B). The mean resting rate of type II neurons (28.9 ± 21.2 spikes/s) was lower than that of the control type II neurons (33.6 ± 20.8 spikes/s) but this difference was not statistically significant (Wilcoxon test, P = 0.75; Fig. 6C).

View larger version (34K): [in this window] [in a new window]   FIG. 5. Properties of the monosynaptically recruited vestibular neurons of the 1-5 h period after a contralateral labyrinthectomy (n = 105). A: histogram of the resting discharges of the neurons. Mean is 41.2 ± 26.9 (SD) spikes/s. There is no silent neuron. Neurons unresponsive to head rotation (), neurons behaving as type I (), and neurons behaving as type II () are considered separately. B: histogram of the sensitivities of the neurons to head rotation. Mean is 0.15 ± 0.25 (SD) spikes/s per deg/s.

View larger version (44K): [in this window] [in a new window]   FIG. 6. Time course of the resting activity in the monosynaptically recruited vestibular neurons after a contralateral labyrinthectomy. For each 4-h recording session, the mean resting rate is represented by a column with an error bar corresponding to the standard deviation of the resting rate. Time indicated under each column refers to the starting time of the corresponding recording session. In A, unresponsive, type I, and type II neurons are pooled. B-D: data related to type I (B), type II (C), and unresponsive (D) subpopulations.

Comparison of the distribution of the sensitivities to horizontal rotation of the neurons encountered just after labyrinthectomy (Fig. 5B) with that of the neurons recruited in the control population (Fig. 4B) clearly shows the global decrease of responsiveness to horizontal rotation caused by the contralateral labyrinthectomy. The ratio of unresponsive neurons is higher after labyrinthectomy (66%). When the sensitivity of the global population was considered, by combining unresponsive units with type I and type II units, the mean sensitivity to head rotation fell from 0.58 ± 0.51 spikes/s per deg/s (control) to 0.15 ± 0.25 spikes/ s per deg/s (Wilcoxon test, P < 0.001; Fig. 7A). The sensitivity of type I units, which was 0.80 ± 0.46 spikes/s per deg/s in the control population, decreased to 0.49 ± 0.26 spikes/ s per deg/s after contralateral labyrinthectomy (P < 0.01) (Fig. 7B). The sensitivity of type II units also decreased from 0.77 ± 0.38 spikes/s per deg/s (control) to 0.31 ± 0.11 spikes/s per deg/s after the lesion (P < 0.001; Fig. 7C).

View larger version (47K): [in this window] [in a new window]   FIG. 7. Time course of the responsiveness to horizontal head rotation of the monosynaptically recruited neurons after a contralateral labyrinthectomy. Data recorded during each 4-h session are lumped and represented by a column the height of which corresponds to the mean value of the studied parameter. Time indicated under the column corresponds to the starting time of the recording session. A: time course of sensitivity to horizontal head velocity of the global neuronal population (unresponsive, type I, and type II neurons are pooled). B: time course of the sensitivity to horizontal head velocity of type I neurons. C: time course of the sensitivity to horizontal head velocity of type II neurons. D: time course of the percentage of unresponsive units in the global population of recorded neurons.

TIME COURSE OF THE POSTCONTRALATERAL LABYRINTHECTOMY CHANGES. Figure 6 presents the mean resting rate of the global neuronal population and of each of the three subpopulations (unresponsive, type I, type II) in the control animals and in the labyrinthectomized animals at different times after the contralateral lesion. This figure shows that there was no shutdown in any of these four populations. During the postlabyrinthectomy studied period, the mean resting rate did not statistically change either in the pooled neuronal population (Kruskal-Wallis test, P = 0.09; Fig. 6A), in the type I neurons (Kruskal-Wallis test, P = 0.33; Fig. 6B), in the type II neurons (Kruskal-Wallis test, P = 0.14; Fig. 6C), or in the unresponsive neurons (Kruskal-Wallis test, P = 0.18; Fig. 6D).

Figure 7 demonstrates that the substantial decrease of responsiveness to head rotation observed just after the contralateral labyrinthectomy persisted, nearly unchanged, during 1 wk. As time went by, the sensitivity to head rotation of the global neuronal population, the type I population, and the type II population remained at a low level (Fig. 7, A-C), whereas the percentage of encountered unresponsive units remained high (Kruskal-Wallis test, P < 0.05; Fig. 7D).

To allow a more detailed comparison of the spiking behavior of the neurons recorded 1 wk after the lesion with that of the neurons recorded 1 h after the lesion, Fig. 8 presents the distribution of the resting rates (Fig. 8A) and of the sensitivities to horizontal head rotation (Fig. 8B) of the neurons recorded 1 wk after the lesion. One week after labyrinthectomy, the mean resting rate of the pooled neuronal population was 48.1 ± 21.8 (SD) spikes/s, those of type I, type II, and unresponsive subpopulations were 56.0 ± 20.3, 42.5 ± 18.5, and 47.2 ± 22.6 spikes/s, respectively. The sensitivity to horizontal head rotation of the global population was 0.20 ± 0.32 spikes/s per deg/s, that of type I neurons was 0.50 ± 0.39 spikes/s per deg/s, and that of type II neurons was 0.29 ± 0.10 spikes/s.

View larger version (36K): [in this window] [in a new window]   FIG. 8. Properties of the monosynaptically recruited vestibular neurons recorded 1 wk after a contralateral labyrinthectomy (n = 90). A: histogram of the resting discharges of the neurons. Mean is 48.1 ± 21.8 (SD) spikes/s. There is no silent neuron. Neurons unresponsive to head rotation (), neurons behaving as type I (), and neurons behaving as type II () are considered separately. B: histogram of the sensitivities of the neurons to head rotation. Mean is 0.20 ± 0.32 spikes/s per deg/s.

BEHAVIOR OF THE POLYSYNAPTICALLY RECRUITED NEURONS AFTER CONTRALATERAL LABYRINTHECTOMY. We identified 208 polysynaptically recruited neurons: 84 in the control population, 36 1 h after the lesion, and between 19 and 27 at the other studied time intervals. The latent period after the electric stimulation ranged between 1.15 and 1.30 ms in 40% of the neurons recorded in intact animals, in 22% of the units obtained 1 h after the lesion, and in 42% of the neurons recorded 1 wk after the lesion.

In the control population, the mean resting rate of the polysynaptically recruited neurons was 29.8 ± 20.4 spikes/s, whereas its mean sensitivity to head rotation was 0.53 ± 0.40 spikes/s.

One hour after the lesion, the mean resting rate of the polysynaptically activated units was larger than that of the control units (49.9 ± 24.8 spikes/s, Wilcoxon test, P < 0.001), whereas their mean sensitivity to horizontal head rotation was lower than that of the control neurons (0.05 ± 0.12 spikes/s per deg/s, Wilcoxon test, P < 0.001).

One week after the lesion, with respect to the control units, the polysynaptically recruited units had a higher mean resting rate (48.2 ± 24.3 spikes/s, Wilcoxon test, P < 0.01) and a lower mean head-velocity sensitivity (0.07 ± 0.15 spikes/s, Wilcoxon test, P <t 0.001).

Neuronal activity after bilateral labyrinthectomy

BEHAVIORAL OBSERVATIONS. Simultaneous bilateral labyrinthectomy resulted in postural instability but not in asymmetrical head and/or body position. Thirty minutes after cessation of halothane delivery, each guinea pig that had undergone a complete bilateral labyrinthectomy recovered a normal level of alertness. The animal could stand up but any attempt to move gave rise to violent rolling toward either side. It had to be restrained gently to avoid hurting itself. This critical instability lasted for 1-2 h. Then the animal did not fall any more, but any movement induced marked oscillations of head and trunk in the horizontal plane. At this stage, the animal could eat food pellets provided on the floor but could not drink without the assistance of someone who provided water in a bottle. One day after the lesion, the animal had a steady gait. Oscillations of head and trunk still were observed but their amplitude was reduced. On the second postoperative day, the animal could drink alone from the bottle. One week after the lesion, the behavior of the animal was roughly normal. However, it walked slowly and avoided brisk movements. Any unexpected change in position caused by the investigator resulted in head and trunk oscillations lasting a few seconds. The vestibuloocular reflex was suppressed completely in each animal during the whole period of study. Figure 9 illustrates this fact by comparing the ocular movement evoked in a same animal by a sinusoidal rotation of the head before (top) and after (middle) bilateral labyrinthectomy.

View larger version (16K): [in this window] [in a new window]   FIG. 9. Vestibuloocular reflex before (top) and after (middle) a bilateral labyrinthectomy in the guinea pig. eh, horizontal eye position; hh, horizontal head position. Vestibuloocular reflex (VOR) was induced by sinusoidal rotation of the animal at 0.3 Hz with a peak angular velocity of 40°/s.

GENERAL CHARACTERISTICS OF THE NEURONS RECORDED AFTER BILATERAL LABYRINTHECTOMY. Recordings were made from 473 monosynaptically recruited neurons in six guinea pigs during 4-h sessions beginning 1 h, 1 day, and 1 wk after a bilateral labyrinthectomy. The spiking parameters of neurons recorded during the same 4-h session were combined.

For each recording session, the distribution of the studied neurons in the explored vestibular areas (S, M, Lv, Ld, and D) is presented in Fig. 10. In each recording session, neurons were studied in each of the explored areas.

View larger version (25K): [in this window] [in a new window]   FIG. 10. Locations of the neurons recorded after bilateral labyrinthectomy. A-C: distribution of the recorded neurons in the S, M, Lv, Ld, and D subdivisions during the different recording periods. In each panel, the time indicated (top right) corresponds to the starting time of the period after the labyrinthectomy during which recordings were made. Used reference map is presented in Fig. 3, A-E.

BEHAVIOR OF THE MONOSYNAPTICALLY RECRUITED NEURONS JUST AFTER BILATERAL LABYRINTHECTOMY. Figure 11 compares for the monosynaptically recruited neurons the distribution of the resting rates of units recorded 1-5 h after a bilateral labyrinthectomy (Fig. 11C) with that of a control neuronal population (Fig. 11A) and that of units recorded 1-5 h after an ipsilateral labyrinthectomy (Fig. 11B) (data from Ris et al. 1995, 1997). From 1 to 5 h after a bilateral labyrinthectomy, 216 monosynaptically recruited neurons were recorded. Their mean resting rate (16.2 ± 22.4 spikes/s) was lower than that of the control population (36.9 ± 21.0 spikes/s, Wilcoxon test, P < 0.001) but higher than that of the neurons recorded 1-5 h after an ipsilateral labyrinthectomy (6.7 ± 17.0 spikes/s, Wilcoxon test, P < 0.001). After a bilateral labyrinthectomy, 53% of the units were silent, whereas there was no silent unit in control animals and a higher percentage of silent units was observed after a unilateral labyrinthectomy (73%).

View larger version (24K): [in this window] [in a new window]   FIG. 11. Comparison of the resting discharges of monosynaptically recruited vestibular neurons recorded in control animals (A), 1-5 h after an ipsilateral labyrinthectomy (data from Ris et al. 1995) (B), and 1-5 h after a bilateral labyrinthectomy (C). Each panel is a histogram of the resting discharges of the neurons recorded in the conditions indicated (top right). Spontaneously active neurons () and silent neurons () are considered separately.

Each of the neurons recorded 1-5 h after a bilateral labyrinthectomy was totally unresponsive to horizontal head rotation. Hence there was no possibility to distinguish type I neurons from type II neurons.

BEHAVIOR OF THE MONOSYNAPTICALLY RECRUITED NEURONS 1 DAY AND 1 WK AFTER BILATERAL LABYRINTHECTOMY. Figure 12 compares the distribution of the resting discharges of neurons recorded 1 day and 1 wk after a bilateral labyrinthectomy (Fig. 12, C and D) with that of neurons recorded 1 day and 1 wk after an ipsilateral labyrinthectomy (Fig. 12, A and B).

View larger version (38K): [in this window] [in a new window]   FIG. 12. Comparison of the resting discharges of monosynaptically recruited vestibular neurons recorded 1 day (A and C) and 1 wk (B and D) after either an ipsilateral (A and B) or a bilateral labyrinthectomy (C and D). Each panel is a histogram of the resting discharges of the neurons recorded in the conditions indicated (top right). Spontaneously active neurons () and silent neurons () are considered separately. Data about ipsilateral labyrinthectomy are from Ris et al. (1997).

Twenty-four hours after a bilateral labyrinthectomy, 174 monosynaptically recruited neurons were recorded. Their mean resting rate (20.8 ± 23.1 spikes/s) was higher than that of the neurons recorded in the same animals 1-5 h after the lesion (Wilcoxon test, P < 0.001). It was not different from that of the neurons recorded 24 h after a unilateral labyrinthectomy (17.6 ± 18.9 spikes/s, Wilcoxon test, P = 0.61) (data from Ris et al. 1997). The percentage of silent neurons was similar 24 h after either a bilateral or an ipsilateral labyrinthectomy (36 and 39%, respectively).

At this time after bilateral labyrinthectomy, the sensitivity to head velocity of each recorded neuron remained nil.

One week after a bilateral labyrinthectomy, 83 monosynaptically recruited neurons were recorded. Their mean resting rate (38.6 ± 21.1 spikes/s) was different neither from that of the control neurons (37.3 ± 20.9 spikes/s, Wilcoxon test, P = 0.48) nor from that of the units recorded 1 wk after an ipsilateral labyrinthectomy (40.8 ± 23.7 spikes/s, Wilcoxon test, P = 0.62) (data from Ris et al. 1995, 1997). As seen 1 wk after an ipsilateral labyrinthectomy, there was no silent unit 1 wk after a bilateral labyrinthectomy.

One week after a bilateral labyrinthectomy, each of the units recorded remained totally insensitive to horizontal head rotation.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The main purpose of this study was to monitor the behavior of a selected population of vestibular neurons after a contralateral or a bilateral labyrinthectomy in the awake guinea pig with an intact CNS. The members of this target neuronal population had to fulfill two criteria: each of them had to be located in the anterior half of the left vestibular nuclear complex and had to be activated by an electric shock applied on the left vestibular nerve. For the units activated at monosynaptic latencies, the major results can be summarized as follows. 1) From 1 h to 1 wk after a contralateral labyrinthectomy, there was a statistically significant increase in the resting activity of type I neurons. 2) Just after a contralateral labyrinthectomy, the sensitivity to horizontal rotation of both the whole target neuronal population and the type I neurons subpopulation was seriously reduced and this decrease remained unchanged during 1 wk. 3) Immediately after a bilateral labyrinthectomy, there was a decrease in the resting discharge of the second-order vestibular neurons. However, the latter was less than that observed after a unilateral ipsilateral labyrinthectomy. 4) After this initial depression, the resting rate of the second-order vestibular neurons totally deprived of their labyrinthine inputs spontaneously recovered. This restoration had a time course similar to that observed after a unilateral ipsilateral labyrinthectomy because it was complete 1 wk after the lesion.

Sampling problems

This study consisted of a comparison of neuronal populations (control and different postoperative times) and therefore had to be designed to avoid possible bias.

The fact that an important anatomic bias has been avoided is attested by the distribution of the locations of the recorded neurons presented in Figs. 3, F-K, and 10, A-C. Neurons were recorded from each of the explored subdivisions during each of the recording sessions. The reference map used in this study to localize the recorded neurons is slightly different from that used in our previous studies on the ipsilateral neurons (Ris et al. 1995). In fact, staining the vestibular nuclear complex for acetylcholinesterase has convinced us that the medial vestibular nucleus extends more caudally than was thought before (Ris et al. 1998). On the other hand, to be more precise, we subdivided here the lateral vestibular nucleus into a dorsal part (Ld) and a ventral part (Lv) because these two regions have different connections (Burian et al. 1990). The distribution of the ipsilateral neurons previously studied by our group within this map can be found in Fig. 2 of Ris et al. (1997).

In studies devoted to the behavior of the vestibular nuclei after a contralateral labyrinthectomy, there are two different search stimuli that can be used to pick up the neurons. In this study, a neuron was selected when it responded to an electrical stimulation of the vestibular nerve, whereas in other studies, a neuron was selected when it responded to horizontal rotation. Both methods (electrical criterion vs. rotation criterion) have advantages and limitations.

Using electrical criteria for selection gives the neurons the same chance to be recorded whatever their static or dynamic characteristics may be but induces a bias in the sampling of type II neurons. Indeed, most of the latter (50% in the cat according to Shimazu and Precht 1966) cannot be activated by an electrical shock on the ipsilateral vestibular nerve. This subpopulation escapes electrical recruitment.

Using rotation criteria allows one to focus attention on type I and type II neurons but introduces another bias. Indeed, our study, which used a selection procedure for the neurons independent of their sensitivity to rotation, has shown that the responsiveness to rotation of a lot of neurons was seriously diminished after a contralateral labyrinthectomy. Consequently, it is logical to think that a number of neurons that were responsive to rotation before the lesion become unresponsive to this stimulation after the contralateral labyrinthectomy and unduly escape the criterion of response to rotation. Using rotation criteria to sample vestibular neurons after a contralateral labyrinthectomy introduces a bias because it favors the most responsive neurons. Moreover, the rotation selection procedure does not make any distinction between the neurons influenced by one labyrinth via a monosynaptic pathway and those activated via a polysynaptic pathway.

In studies devoted to the behavior of the vestibular nuclei after a bilateral labyrinthectomy, only electrical criteria are valid.

Change in resting activity after contralateral labyrinthectomy

McCabe and Ryu (1969) and McCabe et al. (1972) have reported that a unilateral labyrinthectomy is followed by a decrease in resting activity in the vestibular nuclei of both sides. In their studies, the ipsilateral depression was attributed to the withdrawal of the synaptic bombarding from the ipsilateral vestibular nerve, whereas the contralateral depression was ascribed to a compensatory depressive influence from the cerebellum. If right, this phenomenon would have accounted for a dissociation that we found in a previous study (Ris et al. 1997). More precisely, the major part of the recovery from the spontaneous nystagmus was found to occur during the first 10 h after labyrinthectomy while the resting activity of the ipsilateral vestibular neurons remained depressed during the same period of time. However, the present results obtained in awake animals with their cerebellum intact, demonstrate that the "cerebellum shutdown hypothesis" definitively can be ruled out. It is worth noting that the studies of McCabe and Ryu (1969) and McCabe et al. (1972) were not quantitative. They merely described activity as either normal or absent or moderate. Thus our results confirm those of Curthoys et al. (1987), who found a vigorous activity in the vestibular nuclei after a contralateral labyrinthectomy in anesthetized guinea pigs with an intact cerebellum. In the same study, these authors obtained the same result in two animals with a intact cerebellum that were transected spinally and allowed to recover from the surgical anesthesia and maintained under local anesthesia during recording.

In this study, which was performed in awake animals with their CNS intact, it was found that a contralateral labyrinthectomy was followed by an increase in the resting activity of the type I neurons. Comparison with previous work addressing this issue, except that of McCabe and Ryu, shows that the amplitude of this phenomenon, expected from the withdrawal of the inhibitory commissural influence, varied a lot from one study to the other. After a contralateral labyrinthectomy, the resting activity of the type I neurons was found to increase by >100% in the awake cat with its spinal cord transected at C1 (Markham et al. 1977), moderately in the anesthetized guinea pig (Curthoys et al. 1987, 1988; Smith and Curthoys 1988a), only slightly in the decerebrate gerbil (Newlands and Perachio 1990a), and not at all in the decerebrate cat (Shimazu and Precht 1965) and in the anesthetized rat (Hamann and Lannou 1988). It is possible that the sampling bias due to the use of rotation criterion for selection (see Sampling problems) could explain these differences at least in part. Another point that could be of influence resides in the characteristics of the preparation of Markham et al. (1977) that allowed them to record neuronal activity immediately after section of the contralateral vestibular nerve.

Response to head rotation after contralateral labyrinthectomy

Due to the decrease in sensitivity to horizontal rotation induced by contralateral labyrinthectomy, former type I neurons became classified as unresponsive neurons after the lesion (see sampling section). As a result, the sensitivity to rotation of the postlabyrinthectomy type I neurons (Fig. 7B) is overestimated. The sensitivity to rotation of the pooled (unresponsive, type I, type II) neurons (Fig. 7A) thus gives a better assessment of the influence of contralateral labyrinthectomy on the dynamic properties of the second-order neurons. As shown in Fig. 7A, this sensitivity was seriously decreased after a contralateral labyrinthectomy.

A similar decrease in the rotation sensitivity of the type I vestibular neurons immediately after a contralateral labyrinthectomy was found in the decerebrate cat (Shimazu and Precht 1966), in the awake cat with its spinal cord transected at C1 (Markham et al. 1977), and in the decerebrate gerbil (Newlands and Perachio 1990a). In the anesthetized guinea pig tested 52 h after contralateral labyrinthectomy, Smith and Curthoys (1988a) also have reported a statistically significant decrease in gain to 0.2 Hz sinusoidal rotation.

In a previous study, a severe deficit in sensitivity to horizontal rotation was found in the same target neuronal population after an ipsilateral labyrinthectomy (Ris et al. 1995, 1997) (see Table 2). The figures given in Table 2 concerning the sensitivities to head rotation after both an ipsi- and a contralateral labyrinthectomy were all obtained by the multivariate regression analysis (see METHODS). This table shows that just after a labyrinthectomy, the sensitivities to rotation are not only reduced on both sides but are also unequal. It also shows that within 1 wk after the lesion, a rebalancing of the gains occurs. The gains of the ipsi- and the contralateral neurons, albeit seriously reduced, become equal again.

The long-lasting nature of the decrease of contralateral neurons responsiveness observed here already has been described in the decerebrate cat (Precht et al. 1966), in the awake cat with its spinal cord transected at C1 (Yagi and Markham 1984), in the decerebrate gerbil (Newlands and Perachio 1990a), and in the anesthetized guinea pig (Smith and Curthoys 1988a).

Contralateral type II neurons

In fact, after a contralateral labyrinthectomy, the dynamic activity of the vestibular neurons only depends on head velocity signals originating in the intact labyrinth. Therefore, the type II neurons recorded in this study after a contralateral labyrinthectomy owed their behavior to an inhibitory pathway originating in the ipsilateral labyrinth. On the other hand, all the type II neurons recorded after a contralateral labyrinthectomy and the behavior of which is presented in Figs. 6 and 7 were activated at monosynaptic latencies (0.85-1.15 ms) by an electric shock applied on the ipsilateral vestibular nerve. Their mean activation latency was 1.03 ± 0.07 ms 1 h after the lesion and 1.02 ± 0.07 ms 1 wk after the lesion. Among the possible connections that could produce this result, one can surmise a combination of a monosynaptic excitatory input from the ipsilateral otoliths and a disynaptic inhibitory input from the ipsilateral horizontal canal.

Furthermore it has been found that, after contralateral labyrinthectomy, the sensitivity of type II neurons to head rotation was reduced (Fig. 7C). Consistent with this is the fact that in all the previous studies using response to rotation as a search stimulus, type II neurons repeatedly were reported to be very difficult to find (Newlands and Perachio 1990b; Ried et al. 1984; Shimazu and Precht 1965; Smith and Curthoys 1988a). In relative terms, we found a greater ratio of type II neurons than in previous studies. This is probably because of our search stimulus, which after a contralateral labyrinthectomy, favors recruitment of type II neurons.

Comparison with previous studies in bilaterally labyrinthectomized animals

Only two previous studies were devoted to the question of the restoration of the spontaneous activity in the vestibular nuclei after a bilateral labyrinthectomy.

In the first one, conducted in the anesthetized cat, neural activity in the medial vestibular nucleus was found to be depressed 1-2 days after the lesion but normal 1 wk later (Ryu and McCabe 1976). However, this result, obtained only on a qualitative basis, could not be considered as an established fact. The overall resting activity of MVN was only explored by four to six microelectrode penetrations and evaluated as none (electrical silence), normal, or moderate. Moreover, there was no use of any search stimulus. Using a similar procedure, the same authors found a decrease in the activity of the contralateral vestibular nuclei after a unilateral labyrinthectomy (McCabe and Ryu 1969), a result that we could not confirm with our procedure. Nevertheless, as far as the effect of a bilateral labyrinthectomy is concerned, our quantitative results in the awake guinea pig match the qualitative ones obtained by McCabe and Ryu in the anesthetized cat.

In the second study, performed in the awake monkey, an intense spontaneous firing was found in the MVN several months after a section of both vestibular nerves (Waespe et al. 1992). Although there was no comparison with a comparable neuronal population in normal monkeys, this result is in agreement with ours.

Error signal inducing restoration of neuronal activity

After a unilateral labyrinthectomy, the resting discharge of the second-order vestibular neurons in the ipsilateral vestibular nuclei first decreases and then shows spontaneous restoration. The error signal that stimulates this adaptive change is still unknown. Conceivably, it could be any effect of the unilateral labyrinthectomy, from the loss of synaptic drive from the vestibular nerve or the loss of resting activity in the vestibular neurons themselves to the abnormal sensory signals or motor commands that are caused by the symptoms of unilateral labyrinthectomy. In this work, we have demonstrated that the mean resting rate of the second-order vestibular neurons was the same 24 h and 1 wk after either an ipsilateral or a bilateral labyrinthectomy. This proves that the asymmetric symptoms and their associated sensory or motor signals are not the error signal responsible for restoration of neuronal activity after ipsilateral labyrinthectomy because these symptoms are absent after a bilateral labyrinthectomy. Similarly, the asymmetry between the activities in the left and right vestibular nuclei cannot be the error signal.

Source of resting discharge in vestibular neurons

What are the factors responsible for the basal discharge in the monosynaptically activated vestibular neurons and what are the respective weights of these factors? Confronted with these questions, comparison of the resting discharges recorded in two similar populations of monosynaptically activated vestibular neurons after either a bilateral labyrinthectomy (the present work) or an ipsilateral labyrinthectomy (Ris et al. 1995) can be of interest.

Let us consider the very simplified model presented in Fig. 13. Basically, the monosynaptically activated vestibular neurons are influenced by three kinds of signals: a first one originates in the ipsilateral labyrinth and is excitatory (Precht and Shimazu 1965). In Fig. 13, it is represented by x and linked with the representative vestibular neuron by a plus sign. A second one originates in the contralateral labyrinth and is predominantly inhibitory (Furuya et al. 1976; Ried et al. 1984; Shimazu and Precht 1966). In Fig. 13, it is symbolized by y and associated with a minus sign. A third influence is of extralabyrinthine origin. It results from excitatory and inhibitory actions by extravestibular neurons and from some pacemaker activity. In Fig. 13, it is labeled z.

View larger version (10K): [in this window] [in a new window]   FIG. 13. Simple model of the sources of the resting discharge of the monosynaptically activated vestibular neurons. x and y are the ipsilateral labyrinthine input and the contralateral labyrinthine input, respectively. x is associated with a plus sign (excitatory pathway), y with a minus sign (functionally inhibitory commissural pathway). z is an input from extralabyrinthine origin. It includes both influence from nonvestibular neurons and pacemaker activity of the 2nd-order vestibular neurons. It is associated with a plus sign (see text). Percentages associated with x-z correspond to their contribution to the resting rate.

Immediately after a bilateral labyrinthectomy, which nullifies both x and y influences, the mean resting rate of the monosynaptically activated vestibular neurons is equal to ~16 spikes/s. This means that the net effect of z is excitatory. Hence the plus sign associated with z in Fig. 13

(1)

Thus in the intact animal, the recorded mean resting rate of the monosynaptically activated vestibular neurons (37 spikes/s) can be expressed as the result of the following computation where x, y, and z are positive values

(2)

Immediately after an ipsilateral labyrinthectomy, which nullifies the x influence, the mean resting discharge is 7 spikes/s.

Hence

(3)

Solving the system of Eqs. 1-3 gives From this model, the expected mean resting rate after a contralateral labyrinthectomy would be It was actually 41 spikes/s.

The mean resting rate of the monosynaptically activated vestibular neuron is the result of a combination of both positive or negative influences. The percentage of influxes originating from the ipsilateral labyrinth, which is of 53% (30/57), is much higher than that of the influences coming from the contralateral labyrinth (19%). This is in agreement with the results of Ried et al. (1984), who studied the changes in firing rate of the vestibular neurons in response to cathodal or anodal polarizing currents applied either on the ipsi- or on the contralateral labyrinth. They wrote "It is important to mention that the excitatory input from the VIII nerve seemed to be the major source for resting discharge in vestibular neurons, judging by the ease with which they could be silenced by applying anodal stimuli to the ipsilateral labyrinth."

	ACKNOWLEDGEMENTS

  We thank C. Busson for secretarial assistance and realization of the figures and M. Baligniez and B. Foucart for taking care of the mechanical and electronic equipment.

  This research was supported by grants from the Belgian National Fund for Scientific Research, the Belgian Fund for Scientific Medical Research, the Queen Elisabeth Fund for Medical Research, and the Interuniversity Poles of Attraction ProgramBelgian State, Prime Minister's OfficeFederal Office for Scientific, Technical and Cultural Affairs. L. Ris is Research Assistant of the Belgian National Fund for Scientific Research.

	FOOTNOTES

  Address for reprint requests: E. Godaux, Laboratory of Neurosciences, University of Mons-Hainaut, Place du Parc 20, B-7000 Mons, Belgium.

  Received 27 March 1998; accepted in final form 7 July 1998.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AZZENA, G. B. Role of the spinal cord in compensating the effects of hemilabyrinthectomy. Arch. Ital. Biol. 107: 43-53, 1969.[Medline]
• BAARSMA, E. A., COLLEWIJN, H. Changes in compensatory eye movements after unilateral labyrinthectomy in the rabbit. Arch. Otorhinolaryngol. 211: 219-230, 1975.[Medline]
• BURIAN, M., GSTOETTNER, W., MAYR, R. Brainstem projection of the vestibular nerve in the guinea pig: an HRP (horseradish peroxidase) and WGA-HRP (wheat germ agglutinin-HRP) study. J. Comp. Neurol. 293: 165-177, 1990.
• CURTHOYS, I. S., CURTHOYS, E. J., BLANKS, R.H.I., MARKHAM, C. H. The orientation of the semicircular canals in the guinea pig. Acta Otolaryngol. 80: 197-205, 1975.[Medline]
• CURTHOYS, I. S., HARRIS, R. A., SMITH, P. F. The effect of unilateral labyrinthectomy on neural activity in the guinea pig vestibular nuclei. In: The Vestibular System: Neurophysiologic and Clinical Research, edited by M. D. Graham, and J. L. Kemink New York: Raven, 1987, p. 677-687.
• CURTHOYS, I. S., SMITH, P. F., DARLINGTON, C. L. Postural compensation in the guinea pig following unilateral labyrinthectomy. In: Progress in Brain Research. Vestibulo-Spinal Control of Posture and Movement, edited by O. Pompeiano, and J. H. Allum Amsterdam: Elsevier, 1988, vol. 76, p. 375-384.
• DE WAELE, C., GRAF, W., JOSSET, P., VIDAL, P. P. A radiological analysis of the postural syndromes following hemilabyrinthectomy and selective canal and otolith lesions in the guinea pig. Exp. Brain Res. 77: 166-182, 1989.[Medline]
• DE WAELE, C., VIBERT, N., BAUDRIMONT, M., VIDAL, P. P. NMDA receptors contribute to the resting discharge of vestibular neurons in the normal and hemilabyrinthectomized guinea pig. Exp. Brain Res. 81: 125-133, 1990.[Medline]
• DIERINGER, N. "Vestibular compensation." Neural plasticity and its relations to functional recovery after labyrinthine lesions in frogs and other vertebrates. Prog. Neurobiol. 46: 97-129, 1995.[Medline]
• DUENSING, F., SCHAEFER, K. P. Die aktivität einzelner neurone im bereich der vestibulariskerne bei horizontalbeschleunigungen unter besonderer berücksichtigung des vestibulären nystagmus. Arch. Psychiatr. Nervenkr. 198: 225-252, 1958.
• DUTIA, M. B. Vestibular control of neck muscles in acute and chronic hemilabyrinthectomized cats. J. Physiol. (Lond.) 366: 281-290, 1985.[Abstract/Free Full Text]
• FETTER, M., ZEE, D. S. Recovery from unilateral labyrinthectomy in rhesus monkey. J. Neurophysiol. 59: 370-393, 1988.[Abstract/Free Full Text]
• FUCHS, A., ROBINSON, D. A. A method for measuring horizontal and vertical eye movement chronically in the monkey. J. Appl. Physiol. 21: 1068-1070, 1966.[Free Full Text]
• FURUYA, N., KAWANO, K., SHIMAZU, H. Transcerebellar inhibitory interaction between the bilateral vestibular nuclei and its modulation by cerebellocortical activity. Exp. Brain Res. 25: 447-463, 1976.[Medline]
• GSTOETTNER, W., BURIAN, M. Vestibular nuclear complex in the guinea pig: a cytoarchitectonic study and map in three planes. J. Comp. Neurol. 257: 176-188, 1987.[Medline]
• HALMAGYI, G. M., CURTHOYS, I. S., CREMER, P. D., HENDERSON, C. J., TODD, M. J., STAPLES, M. J., D'CRUZ, D. M. The human horizontal vestibulo-ocular reflex in response to high-acceleration stimulation before and after unilateral vestibular neurectomy. Exp. Brain Res. 81: 479-490, 1990.[Medline]
• HAMANN, K. F., LANNOU, J. Dynamic characteristics of vestibular nuclear neurons responses to vestibular and optokinetic stimulation during vestibular compensation in the rat. Acta Otolaryngol. Suppl. 455: 1-19, 1988.
• JENSEN, D. W. Reflex control of acute postural asymmetry and compensatory symmetry after a unilateral vestibular lesion. Neuroscience 4: 1059-1073, 1979.[Medline]
• JUDGE, S. J., RICHMOND, B. J., CHU, F. C. Implantation of magnetic search coils for measurement of eye position: an improved method. Vision Res. 20: 535-538, 1980.[Medline]
• LACOUR, M., XERRI, C. Vestibular compensation: new perspectives. In: Lesion-Induced Neuronal Plasticity in Sensorimotor Systems, edited by H. Florh, and W. Precht Berlin: Springer-Verlag, 1981, p. 240-253.
• LLINAS, R., WALTON, K. Vestibular compensation: a distributed property of the central nervous system. In: Integration in the Nervous System, edited by A. Asanuma, and V. J. Wilson Tokyo: Igaku Shoin, 1979, p. 145-166.
• MAIOLI, C., PRECHT, 1 On the role of vestibulo-ocular reflex plasticity in recovery after unilateral peripheral vestibular lesions. Exp. Brain Res. 59: 267-272, 1985.[Medline]
• MAIOLI, C., PRECHT, W., RIED, S. Short- and long-term modification of vestibulo-ocular response dynamics following unilateral vestibular nerve lesions in the cat. Exp. Brain Res. 50: 259-274, 1983.[Medline]
• MARKHAM, C. H., YAGI, T., CURTHOYS, I. S. The contribution of the contralateral labyrinth to second order vestibular neuronal activity in the cat. Brain Res. 138: 99-109, 1977.[Medline]
• MCCABE, B. F., RYU, J. H. Experiments on vestibular compensation. Laryngoscope 79: 1728-1736, 1969.[Medline]
• MCCABE, B. F., RYU, J. H., SEKITANI, T. Further experiments on vestibular compensation. Laryngoscope 82: 381-396, 1972.[Medline]
• NEWLANDS, S. D., PERACHIO, A. A. Compensation of horizontal canal related activity in the medial vestibular nucleus following unilateral labyrinth ablation in the decerebrate gerbil. I. Type I neurons. Exp. Brain Res. 82: 359-372, 1990a.[Medline]
• NEWLANDS, S. D., PERACHIO, A. A. Compensation of horizontal canal related activity in the medial vestibular nucleus following unilateral labyrinth ablation in the decerebrate gerbil. II. Type II neurons. Exp. Brain Res. 82: 373-383, 1990b.[Medline]
• POMPEIANO, O., XERRI, C., GIANNI, S., MANZONI, D. Central compensation of vestibular deficits. II. Influences of roll tilt on different size lateral vestibular neurons after ipsilateral labyrinth deafferentation. J. Neurophysiol. 52: 18-38, 1984.[Abstract/Free Full Text]
• PRECHT, W., SHIMAZU, H. Functional connections of tonic and kinetic vestibular neurons with primary vestibular afferents. J. Neurophysiol. 28: 1014-1028, 1965.[Free Full Text]
• PRECHT, W., SHIMAZU, H., MARKHAM, C. A mechanism of central compensation of vestibular function following hemilabyrinthectomy. J. Neurophysiol. 29: 996-1010, 1966.[Free Full Text]
• RAPISARDA, C., BACCHELLI, B. The brain of the guinea pig in stereotaxic coordinates. Arch. Sci. Biol. 61: 1-37, 1977.
• RIED, S., MAIOLI, C., PRECHT, W. Vestibular nuclear neuron activity in chronically hemilabyrinthectomized cats. Acta Otolaryngol. 98: 1-13, 1984.[Medline]
• RIS, L., CAPRON, B., DE WAELE, C., VIDAL, P.-P., GODAUX, E. Dissociations between behavioral recovery and restoration of vestibular activity in the unilabyrinthectomized guinea pig. J. Physiol. (Lond.) 500: 509-522, 1997.[Abstract/Free Full Text]
• RIS, L., DE WAELE, C., SERAFIN, M., VIDAL, P.-P., GODAUX, E. Neuronal activity in the ipsilateral vestibular nucleus following unilateral labyrinthectomy in the alert guinea pig. J. Neurophysiol. 74: 2087-2099, 1995.[Abstract/Free Full Text]
• RIS, L., SAUSSEZ, S., GERRITS, N., GODAUX, E., POCHET, R. The subdivisions of the guinea pig vestibular complex revealed by acetylcholinesterase staining. J. Vestib. Res. 8: 1-10, 1998.[Medline]
• RYU, J. H., MCCABE, B. F. Central vestibular compensation. Effects of the bilateral labyrinthectomy on neural activity in the medial vestibular nucleus. Arch. Otolaryngol. 102: 71-76, 1976.[Abstract/Free Full Text]
• SCHAEFER, K.-P., MEYER, D. L. Compensation of vestibular lesions. In: Handbook of Sensory Physiology, edited by and H. H. Kornhuber Berlin: Springer, 1974, p. 463-490.
• SHIMAZU, H., PRECHT, W. Tonic and kinetic responses of cats vestibular neurons to horizontal angular acceleration. J. Neurophysiol. 28: 991-1013, 1965.[Free Full Text]
• SHIMAZU, H., PRECHT, W. Inhibition of central vestibular neurons from the contralateral labyrinth and its mediating pathway. J. Neurophysiol. 29: 467-492, 1966.[Free Full Text]
• SMITH, P. F., CURTHOYS, I. S. Neuronal activity in the contralateral medial vestibular nucleus of the guinea pig following unilateral labyrinthectomy. Brain Res. 444: 295-307, 1988a.[Medline]
• SMITH, P. F., CURTHOYS, I. S. Neuronal activity in the ipsilateral medial vestibular nucleus of the guinea pig following unilateral labyrinthectomy. Brain Res. 444: 308-319, 1988b.[Medline]
• SMITH, P. F., CURTHOYS, I. S. Mechanisms of recovery following unilateral labyrinthectomy: a review. Brain Res. Rev. 14: 155-180, 1989.[Medline]
• SMITH, P. F., DARLINGTON, C. L., CURTHOYS, I. S. The effect of visual deprivation on vestibular compensation in the guinea pig. Brain Res. 364: 195-198, 1986.[Medline]
• VIBERT, N., DE WAELE, C., ESCUDERO, M., VIDAL, P. P. The horizontal vestibulo-ocular reflex in the hemilabyrinthectomized guinea-pig. Exp. Brain Res. 97: 263-273, 1993.[Medline]
• WAESPE, W., SCHWARZ, U., WOLFENSBERGER, M. Firing characteristics of vestibular nuclei neurons in the alert monkey after bilateral vestibular neurectomy. Exp. Brain Res. 89: 311-322, 1992.[Medline]
• XERRI, C., GIANNI, S., MANZONI, D., POMPEIANO, O. Compensation of central vestibular deficits. I. Response characteristics of lateral vestibular neurons to roll tilt after ipsilateral labyrinth deafferentation. J. Neurophysiol. 50: 428-448, 1983.[Free Full Text]
• YAGI, T., MARKHAM, C. H. Neural correlates of compensation after hemilabyrinthectomy. Exp. Neurol. 84: 98-108, 1984.[Medline]
• ZENNOU-AZOGUI, Y., BOREL, L., LACOUR, M., EZ-ZAHER, L., OUAKNINE, M. Recovery of head postural control following unilateral vestibular neurectomy in the cat. Acta Otolaryngol. Suppl. 509: 3-19, 1993.

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society