JN Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Neurophysiol 79: 2690-2715, 1998;
0022-3077/98 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wearne, S.
Right arrow Articles by Cohen, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wearne, S.
Right arrow Articles by Cohen, B.

The Journal of Neurophysiology Vol. 79 No. 5 May 1998, pp. 2690-2715
Copyright ©1998 by the American Physiological Society

Control of Spatial Orientation of the Angular Vestibuloocular Reflex by the Nodulus and Uvula

Susan Wearne, Theodore Raphan, and Bernard Cohen

Departments of Neurology and Biophysics and Physiology, Mount Sinai School of Medicine, New York 10029; and Computer and Information Sciences, Brooklyn College of the City University of New York, Brooklyn, New York 11210

    ABSTRACT
Abstract
Introduction
Methods
Results
Discussion
References

Wearne, Susan, Theodore Raphan, and Bernard Cohen. Control of spatial orientation of the angular vestibuloocular reflex by the nodulus and uvula. J. Neurophysiol. 79: 2690-2715, 1998. Spatial orientation of the angular vestibuloocular reflex (aVOR) was studied in rhesus monkeys after complete and partial ablation of the nodulus and ventral uvula. Horizontal, vertical, and torsional components of slow phases of nystagmus were analyzed to determine the axes of eye rotation, the time constants (Tcs) of velocity storage, and its orientation vectors. The gravito-inertial acceleration vector (GIA) was tilted relative to the head during optokinetic afternystagmus (OKAN), centrifugation, and reorientation of the head during postrotatory nystagmus. When the GIA was tilted relative to the head in normal animals, horizontal Tcs decreased, vertical and/or roll time constants (Tcvert/roll) lengthened according to the orientation of the GIA, and vertical and/or roll eye velocity components appeared (cross-coupling). This shifted the axis of eye rotation toward alignment with the tilted GIA. Horizontal and vertical/roll Tcs varied inversely, with Tchor being longest and Tcvert/roll shortest when monkeys were upright, and the reverse when stimuli were around the vertical or roll axes. Vertical or roll Tcs were longest when the axes of eye rotation were aligned with the spatial vertical, respectively. After complete nodulo-uvulectomy, Tchor became longer, and periodic alternating nystagmus (PAN) developed in darkness. Tchor could not be shortened in any of paradigms tested. In addition, yaw-to-vertical/roll cross-coupling was lost, and the axes of eye rotation remained fixed during nystagmus, regardless of the tilt of the GIA with respect to the head. After central portions of the nodulus and uvula were ablated, leaving lateral portions of the nodulus intact, yaw-to-vertical/roll cross-coupling and control of Tcvert/roll was lost or greatly reduced. However, control of Tchor was maintained, and Tchor continued to vary as a function of the tilted GIA. Despite this, the eye velocity vector remained aligned with the head during yaw axis stimulation after partial nodulo-uvulectomy, regardless of GIA orientation to the head. The data were related to a three-dimensional model of the aVOR, which simulated the experimental results. The model provides a basis for understanding how the nodulus and uvula control processing within the vestibular nuclei responsible for spatial orientation of the aVOR. We conclude that the three-dimensional dynamics of the velocity storage system are determined in the nodulus and ventral uvula. We propose that the horizontal and vertical/roll Tcs are separately controlled in the nodulus and uvula with the dynamic characteristics of vertical/roll components modulated in central portions and the horizontal components laterally, presumably in a semicircular canal-based coordinate frame.

    INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References

The angular vestibuloocular reflex (aVOR) exhibits the property of spatial orientation. During nystagmus induced by rotation of the subject or surround, the yaw axis component of slow-phase velocity tends to align with gravity or with the gravito-inertial acceleration vector (GIA) (Angelaki and Hess 1994a,b; Dai et al. 1991; Harris 1987; Harris and Barnes 1987; Raphan and Cohen 1986; Raphan et al. 1992, 1996; Raphan and Sturm 1991; Wearne et al. 1996, 1997a,b). Alignment of the eye velocity vector with the GIA has been modeled by three processes: a reduction in the time constant of the dominant horizontal component, an increase in the torsional and/or vertical time constants, and the appearance of orthogonal vertical or torsional "cross-coupled" eye velocity (Dai et al. 1991; Raphan and Cohen 1996; Raphan et al. 1992; Raphan and Sturm 1991).1 We have shown that the velocity storage mechanism in the vestibular system is responsible for spatial orientation of the aVOR, and that the direct visual and vestibular pathways do not orient eye velocity to the GIA (Wearne et al. 1997a). The study of the spatial orientation of the aVOR is therefore essentially a study of the three-dimensional characteristics of velocity storage. Because velocity storage is activated by somatosensory input in both monkey (Solomon and Cohen 1992) and humans (Bles et al. 1984), this type of spatial orientation is likely to be important for postural and gaze stabilization during angular locomotion (Cohen 1998).

Behavioral, lesion, and stimulation studies indicate that the nodulus and uvula control the horizontal time constant of velocity storage, one element in determining its spatial orientation. The horizontal aVOR time constant is reduced when the head is reoriented relative to gravity during postrotatory nystagmus (tilt dumping), both in humans (Benson 1974; Fetter et al. 1992) and in monkeys (Angelaki and Hess 1994a; Merfeld 1995; Merfeld et al. 1993; Raphan et al. 1981; Solomon and Cohen 1994; Waespe et al. 1985). The horizontal time constant is also reduced when subjects view a relative stationary visual surround during vestibular nystagmus or optokinetic afternystagmus (OKAN; light dumping) (Cohen et al. 1977, 1981; Waespe et al. 1983, 1985). Electrical stimulation of the nodulus and lobule 9d of the ventral uvula reproduces this reduction (Solomon and Cohen 1994). Both tilt dumping and light dumping are lost after ablation of the nodulus and uvula (Angelaki and Hess 1994a, 1995; Waespe et al. 1985).

The nodulus and ventral uvula are likely also to be involved in producing the changes in vertical and/or torsional components that tilt the eye velocity axis during spatial orientation. The cortex of the nodulus and uvula has been divided into parasagittal zones comprising strips of Purkinje cells, innervated by subnuclei of the inferior olive (Voogd 1964, 1969; Voogd et al. 1996; Voogd and Bigaré 1980). There are both visual (optokinetic) and vestibular climbing fiber afferents to the nodulus of the rabbit, and their sensitivity axes are aligned approximately with the axis of one of the three semicircular canals (Barmack and Shojaku 1992, 1995; Kano et al. 1990; Wylie et al. 1994). (The sensitivity axis of a neuron is that axis around which rotation causes a maximal modulation of its firing rate.) These axes, derived from physiological studies in the rabbit, are shown projected onto a monkey head in Fig. 1B. Activity induced by yaw optokinetic stimulation and nystagmus (OKN) about a vertical axis [VA (Yaw) OKN] is processed in the caudal dorsal cap (cdc), whereas activity related to optokinetic stimulation about an axis parallel to that of the ipsilateral anterior canal [horizontal axis, HA (135°) OKN] reaches the cerebellar cortex through the rostral dorsal cap (rdc) and ventrolateral outgrowth (vlo) (Balaban and Henry 1988; Barmack and Shojaku 1992; Graf et al. 1988; Kano et al. 1990; Katayama and Nisimaru 1988; Leonard et al. 1988; Simpson et al. 1981; Takeda and Maekawa 1984). Vestibular signals aligned approximately with the anterior and posterior canal axes [AC and PC (135°) VOR signals] project from the beta nucleus of the inferior olive (Alley et al. 1975; Balaban and Henry 1988; Barmack et al. 1993a; Barmack and Shojaku 1992, 1995; Katayama and Nisimaru 1988; Shojaku et al. 1991). Utricular signals project from the dorsomedial cell column (DMCC) (Barmack and Fagerson 1994) and the beta nucleus (Barmack et al. 1989, 1993b). As yet, no lateral canal-related signals have been found in the inferior olive or in climbing fiber responses (Barmack and Shojaku 1992, 1995).


View larger version (35K):
[in this window]
[in a new window]
  		FIG. 1. A: right-handed coordinate frame to which components of the eye velocity vector are referred. Positive direction of rotation about each axis is defined according to the right hand rule, as leftward yaw (+omega Z), downward pitch (+omega X) and counterclockwise roll (+omega Y), from the animal's point of view. B: relation between the head reference frame used to describe 3-dimensional eye movements and maximal sensitivity axes for visual modulation of neurons in the left nodulus and right inferior olive of the rabbit (modified from Simpson and Graf 1985). The excitatory direction for visually and vestibularly induced complex spike responses of Purkinje cells in the left nodulus of the rabbit is indicated by the curved arrows about axes VA and HA (135°) (Barmack and Shojaku 1992). C-E: paradigms used to tilt the gravito-inertial acceleration vector (GIA) relative to the head during optokinetic or vestibular nystagmus. C: continuous tilts of GIA induced during optokinetic afternystagmus (OKAN) with the head upright and tilted left ear down (LED). Ag is the equivalent acceleration due to gravity. The angular velocity of the surround, along the Z-axis of the head, is given by omega OZ. D: reorientation of the head with respect to the GIA during postrotatory nystagmus with the GIA equivalent to the vector of gravitational acceleration (Ag). The angular velocity of the head along the Z-axis is given by omega HZ. During postrotatory nystagmus, the animal is tilted in the roll plane by the angle theta , shifting the GIA relative to the head. E: centrifugation facing motion, centered and back to motion. Centripetal acceleration is indicated as Ac. The vector sum of Ag and Ac is the GIA vector. The angle between Ag and the GIA in the roll plane is given as theta R. For +omega HZ rotation, the animal was oriented right ear out when facing motion, and left ear out (LEO) when back to motion. An analogous stimulus paradigm was given for -omega HZ rotation, with left ear out when facing motion and right ear out (REO) when back to motion.

Unlike the rabbit, no physiological studies relating zonal organization to function have been performed in rhesus monkey. Anatomic studies, however, have demonstrated a similar zonal organization in the rhesus monkey to that of the rabbit (Voogd et al. 1996). Four sagittal zones have been distinguished, based on four white matter compartments innervated by discrete subnuclei of the contralateral inferior olive. The most medial zone (Zone 1) extends from the midline to 0.8 mm lateral in the nodulus and uvula. It receives composite climbing fiber input from the beta nucleus medially and the lateral caudal dorsal cap. If it is assumed that the same canal-, otolith-, and OKN-related information is present in the inferior olive of the monkey as in the rabbit, Zone 1 has horizontal axis [vertical canal, HA (135°) VOR] and otolith (beta ) as well as yaw axis representation [cdc; VA (Yaw) OKN]. The second zone (Zone 2) extends from 0.8 to 2.4 mm in the nodulus and uvula, and forms the lateral limits of the vermis over the posterior surface of the cerebellum. Zone 2 was labeled from injections that included the ventrolateral outgrowth, rostral beta, and rostral medial accessory olive (rMAO) in rhesus monkey (Voogd et al. 1996). Input to this zone from vlo and from rostral beta  is related in the rabbit to HA (135°) OKN, HA (135°) VOR, and the utricle. Zone 3, innervated by the caudal dorsal cap in the monkey, would receive information about yaw axis OKN [VA (Yaw) OKN]. It extends from 2.4 to 3.2 mm in the nodulus and is restricted to the nodulus. Zone 4 extends from 3.2 to 4 mm lateral. Its olivocerebellar afferents are currently unidentified in rhesus monkey, but by analogy with the lateral zone in rabbit, probably include the DMCC and rMAO (Balaban and Henry 1988; Katayama and Nisimaru 1988; Tan et al. 1995). We have adopted the classification of Voogd et al. (1996) for the animals in our series, and these four zones are represented in Figs. 4 and 18.


View larger version (69K):
[in this window]
[in a new window]
  		FIG. 4. Schematic diagram indicating complete and partial lesions of nodulus and uvula in the monkeys of this series, drawn on sagittal zones defined by Voogd et al. (1996) and J. Voogd (personal communication). Zones are based on AChE staining of margins of white matter compartments, which carry the climbing fibers to each zone. Numbers below represent mm from the midline.


View larger version (25K):
[in this window]
[in a new window]
  		FIG. 18. Model of 3-dimensional control of orientation and dynamics of velocity storage system by the canal related zones of the nodulus and ventral uvula. Above are shown the 4 zones of the rhesus monkey nodulus, based on inferior olive input (Voogd et al. 1996). Zone 1 receives input from the beta  nucleus and the caudal dorsal cap (cdc). Zone 2 receives input from the ventrolateral outgrowth (vlo), and Zone 3 from cdc. Inputs to Zone 4 are still unidentified in the monkey. The inputs to the various subnuclei of the inferior olive are shown above and on left. G0, gain matrix in integrator pathway; G1, direct vestibular pathway gain matrix; Tcan, matrix of head to canal coordinate transformation. See text for details.

It has been postulated that an important function of the nodulus and ventral uvula of the rabbit is to determine three-dimensional spatial orientation (Barmack and Shojaku 1992). The same is likely true for the monkey. Accordingly, complete or partial nodulo-uvulectomy was reported to impair the dynamics of the low-frequency torsional aVOR in the monkey, leaving the vertical aVOR relatively unaffected (Angelaki and Hess 1994a, 1995). Preliminary results following similar nodulo-uvular lesions, however, suggest that the nodulus and uvula control aVOR dynamics in all dimensions (Wearne et al. 1996).

The purpose of this study was to characterize the three-dimensional behavior of eye velocity during paradigms that shift the GIA relative to the head before and after partial and complete lesions of the nodulus and uvula. Constant velocity centrifugation provides a robust technique for shifting the GIA vector relative to the head, and reorientation of eye velocity axes to the GIA has not been studied after nodulo-uvulectomy with centrifugation. In addition to centrifugation, we utilized previously studied paradigms, such as OKAN in tilted positions, tilt dumping, and light dumping. We wished to determine whether the nodulus and uvula control spatial orientation of the aVOR around all axes. In the course of this study, it became apparent that processes related to control of time constants around different axes could be separated and localized to different regions of the nodulus and uvula. A preliminary report has been presented (Wearne et al. 1996).

    METHODS
Abstract
Introduction
Methods
Results
Discussion
References

Juvenile rhesus monkeys (Macaca mulatta; weight 3-4 kg) were used in these studies. Quantitative data came largely from two animals with representative complete and partial lesions (M9312 and M9303), but relevant results from five other monkeys with complete (M1107, M1152, and M1153) and partial lesions (M1173 and M1175) from previous studies (Cohen et al. 1992; Waespe et al. 1985) were reanalyzed to expand the database. The experiments conformed to the Principles of Laboratory Animal Care (NIH Publication 85-23, Revised 1985), and were approved by the Institutional Animal Care and Use Committee.

Surgical procedures

Animals were prepared initially under anesthesia with a frontal plane coil to record horizontal and vertical eye position (Judge et al. 1980; Robinson 1963). A roll coil was threaded around the superior rectus muscle on top of the eye to record roll eye position (Dai et al. 1994; Yakushin et al. 1995). Bolts implanted on the skull provided painless head stabilization during experiments. The animals were allowed to recover from anesthesia and received baseline testing. At a later date, they were reanesthetized, and a 25% solution of mannitol (250 mg/ml) was administered to a dose of 1 gm/kg to reduce brain vascularity and edema. Under gas inhalation anesthesia, the dura and pia mater were opened, and the cerebellum was exposed. The secondary fissure was identified, and a horizontal dissection plane was established below the secondary fissure, between lobules 9a and 9b. The rostral extension of this plane led to the fourth ventricle just caudal to the fastigial nucleus. Partial lesions were confined to the central vermis. The nodulus (lobule 10) and ventral uvula (sublobules 9c and 9d) were aspirated from the midline to the lateral margins of the uvula (approximately ±2 mm). Complete nodulo-uvulectomy was performed by extending the dissection laterally in its rostral portions to include the lateral 2 mm of the nodulus (Madigan and Carpenter 1971; Snider and Lee 1961). More of the dorsal uvula (sublobules 9a and 9b) was generally removed in the complete than the partially lesioned animals. Bleeding was generally slight and was controlled with cautery. Vomiting, which can occur postoperatively after nodulo-uvulectomy (Cohen et al. 1992; Waespe et al. 1985), was prevented by prophylactic administration of promethazine HCl(2 mg/kg). Animals also received analgesics and antibiotics postoperatively. The animals were initially posturally unstable, but this gradually disappeared over the next several weeks.

Stimulation apparatus

Testing began ~1 wk after operation and continued for 12-36 mo. Eye movements were induced by optokinetic stimulation, by angular acceleration about an earth vertical axis with the animal either centered or at a 25-cm radius from the axis of rotation, or by rapidly reorienting the animal during postrotatory vestibular nystagmus. The stimulator (Neurokinetics, Pittsburgh, PA), has been described previously (Wearne et al. 1996, 1997a,b). In brief, it had three independently controlled, gimbaled axes of rotation: an outer horizontal axis, a nested yaw axis, and a doubly nested inner pitch-roll axis. The yaw and pitch-roll axes were enclosed in a light-tight, optokinetic sphere, 109-cm diam, with 10° vertical black and white stripes on the inside. The axis of the OKN sphere, which was also independently controlled, was collinear with the yaw axis. When the OKN sphere rotated, it produced full field motion that induced OKN and OKAN.

Monkeys were seated in a primate chair with the head and eyes centered in the field coils. The animals could be positioned either with the head centered with respect to each rotation axis or at the end of the centrifuge arm, directing centripetal acceleration along the interaural or the nasooccipital axes. With the monkey erect, the yaw axis was aligned with gravity, and the horizontal stereotaxic plane was aligned with the gravitational horizontal. Thus the lateral semicircular canals were tilted up ~30° from the earth horizontal plane during the experiments (Blanks et al. 1985; Yakushin et al. 1995).

Coordinates and notation

The three-dimensional components of the vestibular stimuli and oculomotor responses were referenced to the right-handed Cartesian coordinate system shown in Fig. 1A. The X-, Y-, and Z-axes denote the interaural (positive left), nasooccipital (positive backward), and dorsoventral (positive up) axes of the head, respectively. Three-dimensional eye orientation is represented as Euler angles using the Fick rotation convention.2 The symbols, phi , theta , and psi , represent rotations about the Z-axis, the rotated X-axis, and the doubly rotated Y-axis (visual axis). These angular deviations from the reference position will be referred to as horizontal, vertical, and roll eye positions, respectively. The components of the eye velocity vector in head coordinates, symbolized by omega  = [omega X, omega Y, omega Z] that denoted vertical, torsional, and horizontal components, were obtained directly from the Euler angles and their derivatives (Goldstein 1980; Yakushin et al. 1995). Positive directions for the rotations are in accordance with the right-hand rule (Fig. 1A).

To provide a basis for comparison, the coordinate frame determined from studies in the rabbit (Barmack et al. 1993b; Barmack and Shojaku 1992; Wylie et al. 1994) has been superimposed on the coordinate frame utilized in this study on the monkey (Fig. 1B). The axes of the ipsilateral anterior and posterior canals (AC and PC), are oriented at ~45 and 135° from the nasooccipital axis of the head. The vertical axis (VA) is aligned with the +Z head axis, whereas the positive pole of the horizontal axis (HA) (135°) is rotated counterclockwise about the +Z-axis by 135° from the forward-pointing pole of the nasooccipital (-Y) axis. The VA and HA axes are those axes about which rotation of the head or visual surround maximally activates climbing fibers and complex spike responses of Purkinje cells (Barmack and Shojaku 1992; Kano et al. 1990; Simpson and Alley 1974; Wylie et al. 1994). Thus the VA axis aligns approximately with the axis of the lateral canals, and the HA (135°) axis aligns with the axis of the ipsilateral (left) anterior canal and contralateral (right) posterior canal.

Calibration of eye position

Eye orientation was measured using the implanted scleral search coils. The normal to the frontal coil on the implanted eye was aligned with the visual axis. Portions of the frontal coil voltages were subtracted from the roll coil voltages to remove cross-talk as the upright animal was rotated around a spatial vertical axis. To calibrate eye position in untrained monkeys, we made use of the fact that Listing's plane is aligned approximately with the coronal head plane and is independent of the initial reference position from which eye position axes are calculated (Helmholtz 1866; Tweed and Vilis 1990). We assumed that the average coil voltages during spontaneous positions of fixation while the animal made saccades for 15-30 s in light correspond to the straight ahead eye position with zero roll. This assumption is consistent with data obtained in trained and calibrated monkeys for a large range of saccades in light and dark (van Opstal et al. 1995). When the eye is in this orientation, the visual axis is aligned with the -eY axis of the head frame. To a good approximation, this aligns with the stereotaxic coordinate frame of the head, which we physically aligned with the axes of the horizontal and vertical field coils. More complete details of the calibration of eye orientation are presented in Yakushin et al. (1995).

Experimental protocol

The order of testing was independently randomized for each monkey, and repeated measures were performed on each experimental condition. Amphetamine sulfate (0.3 mg/kg) was given 30 min before testing to maintain alertness. Monkeys were rotated in yaw to habituate the time constant of velocity storage to some stable value. This ensured that effects of habituation during the experiments would not significantly influence the experimental results (Cohen et al. 1992). As a result the time constant and initial value of OKAN for the upright position remained approximately constant throughout the period of testing. The horizontal OKAN time constant in the upright position was ~15-25 s, depending on the monkey. Intrasubject variability remained low throughout the period of testing.

Eye movements were induced in three stimulus paradigms. The animals were given steps of yaw axis optokinetic stimulation at 60°/s and 90°/s, lasting 30 s (Fig. 1C). This induced horizontal OKN and OKAN. OKN was elicited in light and OKAN in darkness. Tilts were left or right ear down (LED or RED) with regard to gravity at angles of 0, 17, 36, 45, 52, and 90°. The GIA was tilted by eccentric constant velocity rotation on a centrifuge in darkness, either facing the direction of motion (Fig. 1E, left), with back to the motion (Fig. 1E, right) or facing radially, nose-in or nose-out (Fig. 11B). Centrifugation was called tangential when facing or back-to-motion and radial when nose-in or nose-out. The centrifuge was accelerated at 40°/s2 to a final angular velocity of 400°/s. This generated a maximum interaural centripetal acceleration of 1.24 g and tilted the GIA vector, the sum of gravitational (Ag) and centripetal (Ac) accelerations, by 51° with respect to gravity. The final stimulus velocity was maintained for at least 120 s. For +Z rotation, animals were right ear out (REO) when facing motion (Fig. 1E, left) and left ear out (LEO) when back to motion (Fig. 1E, right). For -Z rotation, these conventions were reversed. For facing and back to motion centrifugation, Ac was directed along the interaural axis, and the GIA tilted dynamically in the roll plane of the head, through an angle (theta ) that increased with the angular velocity of the centrifuge. Animals were also rotated about a centered axis (Fig. 1E, middle).


View larger version (31K):
[in this window]
[in a new window]
  		FIG. 11. Two-dimensional phase plane trajectories of eye rotation axes during centrifugation in M9312, which tilted the GIA by 52° in the roll (A) or pitch plane (B). Each graph is composed of 4 trials, one for +Z and another for -Z rotation before (black-square) and after operation (open circle ). Circles start at the points farthest from the origin on the top and bottom halves of each graph and proceed toward zero. A: eye rotation axis during facing and back to motion centrifugation with right ear out (left panel), centered (middle panel), and left ear out (right panel). Preoperatively, the eye velocity axis gradually aligned with the GIA. Postoperatively, axis reorientation was abolished. B: eye rotation axis during centrifugation with nose out (left panel), centered, and nose in (right panel). Preoperatively, the axis tilted in the animal's pitch plane toward alignment with the GIA. Postoperatively, axis reorientation was abolished.

In addition, animals were tilted with respect to the GIA by reorienting them about the roll or pitch axes during postrotatory nystagmus in darkness (tilt dumping). Nystagmus was induced by rotation at 60°/s about a spatial vertical axis. When the per-rotatory nystagmus had decayed to zero, rotation was stopped, inducing postrotatory nystagmus. Two seconds later, the head was rapidly reoriented by rotating the animal about a spatial horizontal axis (Fig. 1D, right). In the normal monkey, this causes a reduction in the horizontal time constant (tilt dumping) and the appearance of concurrent vertical and/or torsional velocities (cross-coupling).

The ratio of the recovery velocity to the velocity just before exposure to the light is a measure of the effects of visual suppression on the velocity storage integrator (Cohen et al. 1977) and was compared in the normal, completely lesioned and partially lesioned animals. These results are reported in Table 1 in which dumping was considered absent if the recovery velocity reached levels within 20% of those expected as a result of decline of eye velocity in darkness, plus losses due to visual feedback, determined from the rising time constant (Raphan et al. 1979; Waespe et al. 1983, 1985). If the velocity profile was reduced by >20% from the expected level, then it was labeled as positive (+). For tilt dumping, the plus sign in Table 1 indicates that the horizontal velocity storage time constant was reduced by tilt dumps, and the minus sign indicates a lack of effect. Dumping was considered absent in response to tilt (-) if the time constant of decline in horizontal slow-phase velocity was consistently within 30% of the time constant recorded in the upright position.

 
View this table:
[in this window] [in a new window]
  		TABLE 1. Horizontal and vertical VOR gains and time constants for nodulectomy animals

Data acquisition and processing

Eye position and photocell data were sampled at a rate of 600 Hz per channel. Before sampling, the eye position data were prefiltered by an 8-pole Butterworth filter with a cutoff corner frequency of 30 Hz. Slow-phase eye velocity was obtained by calculating the vector of eye velocity in head coordinates and removing saccades, using an order statistic filter (Engelken and Stevens 1990, 1991). Slow-phase velocities were analyzed from the onset of OKAN to the point where the yaw axis eye velocity decayed to zero. OKAN time constants were estimated by fitting a single exponential function to the decaying portion of eye velocity starting 2 s after the end of stimulation (Cohen et al. 1977). During steps of angular velocity for M9312 and M9303, velocity storage time constants were estimated by fitting a sum of two exponential functions to the decaying portion of eye velocity (Raphan et al. 1979). The cupula time constant was constrained to 4 s (Goldberg and Fernández 1971), and the initial integrator state was constrained to be zero. Whenever a single time constant is quoted for these animals, it is the central or velocity storage time constant. If secondary nystagmus appeared during measurement of aVOR time constants, the offset due to overshoot was subtracted before time constants were fitted. Time constants for the other animals were determined from paper records by dividing the area under the slow-phase velocity envelope by the initial jump in velocity (Raphan et al. 1979). Animals in whom a single area time constant estimate was used are marked with double asterisks in Table 1. The direction of nystagmus is denoted in the figures and the text by the direction of the slow-phase velocity.

Slow-phase velocities during centrifugation were analyzed from the end of the angular acceleration until the nystagmus had decayed to zero. Slow-phase velocity was related to the eigenvalues and eigenvectors of the velocity storage system matrix (Raphan and Sturm 1991) by fitting the best tangent line to the three-dimensional eye velocity vector starting 10 s after reaching constant angular head velocity (Wearne et al. 1997a,b). Because the decay of the eye velocity trajectory approaches the yaw axis eigenvector in the limit (Raphan et al. 1992; Raphan and Sturm 1991), this tangent line is a close approximation to the eigenvector. The magnitude of axis shift in the head coordinate frame was computed as the angular difference between the yaw eigenvector and the Z-axis of the head. The three-dimensional tangent line was then projected into each cardinal head plane, and the angle made by this projection and the Z-axis was computed in each plane. Goodness of fit was verified in each data trace by plotting the fitted eigenvector over the components of desaccaded eye velocity trajectories in two and three dimensions (Wearne et al. 1996, 1997a,b). In the text, single values are used when describing specific data traces to aid the reader in evaluating the figures. Whenever a data set was considered, means ± SD are given.

At the end of the experiments, animals were deeply anesthetized and perfused intracardially with saline and 10% Formalin. The brains were removed, embedded in celloidin, and serially sectioned in sagittal stereotaxic planes. Sections were stained with cresyl violet, and the extent of the lesions was reconstructed. The atlas of Madigan and Carpenter (1971) was used as an aid in plotting the location of the lesions. Roman numerals were used to denote the uvula (IX) and nodulus (X) for clarity. In other contexts, the numerals 9 and 10 refer to the uvula and nodulus, respectively.

    RESULTS
Abstract
Introduction
Methods
Results
Discussion
References

Description of nodulo-uvular lesions

Sagittal sections in monkey M9312, in whom the entire nodulus and most of the uvula were removed are shown in Fig. 2, A and B. The lesion was also reconstructed on sagittal sections from a normal rhesus cerebellum (Fig. 2C) (Madigan and Carpenter 1971). The sections, ~0.5, 2.5, and 3.75 mm to the left (A) and right (B) of the midline, correspond to Zones 1, 3, and 4 of Voogd et al. (1996) in the rhesus monkey. Medial portions of the nodulus and uvula were completely removed, and tissue was absent ventral to the secondary fissure (2°) [0.5 mm, Fig. 2, A, (1)B, (1) and C, Zone 1]. Laterally, the nodulus and uvula were completely ablated on the right (Fig. 2B, (3) R 2.5 mm; Fig. 2C, Zone 3, 2.5 mm), but a small portion of the dorsal uvula remained 2.5 mm from the midline on the left (Fig. 2A, (3), L 2.5 mm). The uvula was absent at 3.75 mm level, and the paramedian lobules (PML) were shifted toward the midline [Figs. 2A, (4) and B, (4), 3.75 mm]. Three other animals (M1107, M1152, and M1153) from an earlier study (Waespe et al. 1985) had similarly complete lesions (Table 1).


View larger version (68K):
[in this window]
[in a new window]
  		FIG. 2. Sagittal sections through the cerebellum in M9312 in whom the nodulus and most of the uvula were removed. The drawings correspond to sections at ~0.5 mm, 2.5 mm, and 3.75 mm from the midline to the left (A) and right (B) of the midline. The numbers 1, 3, and 4 correspond to the rhesus monkey zones of Voogd et al. (1996). The arrows point to the paramedian lobule in the lateral-most sections. C: normal rhesus monkey sagittal sections (Madigan and Carpenter 1971) at 0.5 mm (Zone 1), 2.5 mm (Zone 3), and 3.75 mm (Zone 4) from the midline. The superimposed shaded regions indicate the extent of the lesion in M9312. Because the lesions were symmetric, only one side is shown in the reconstruction. FN, fastigial nucleus; AIN, anterior interposed nucleus; PIN, posterior interposed nucleus; PML, paramedian lobule. 1° and 2° refer to the primary and secondary fissures. IX and X refer to the nodulus and uvula, respectively. The uvula is further divided into sublobulesa-d.

Partial lesions of the nodulus and uvula were made in M9303 in the current study and in M1173 and M1175 in an earlier study (Cohen et al. 1992). In each case the lesions were confined to the vermal portions of the nodulus and uvula, and the lateral portions of the nodulus were intact. The lesion was symmetric in M9303. Sections of the lesion 0.5, 2.5, and 3.75 mm from the midline in M9303 are drawn in Fig. 3, A and B, and the extent of the lesion was reconstructed on a normal cerebellum in Fig. 3C. The lesions in M1173 and M1175 are presented in Fig. 9 of Cohen et al. (1992), and a section is shown in Fig. 14B. In M9303, the central regions of lobules 9d and 10 of the caudal vermis were removed en bloc, extending to the lateral margins of the uvula [Fig. 3A, (1) B, (1) and C, Zone 1 (0.5 mm)]. The lateral-most 1.8 mm of the nodulus, corresponding to rhesus monkey zones 3 and 4 of Voogd et al. (1996), were left intact [Fig. 3, A-C (2.5 and 3.75 mm)]. Purkinje cells in this region had a normal appearance with intact nuclei and cytoplasm. Approximately 70% of the nodulus and rostral-ventral uvula (sublobule 9d) were destroyed in M1173 and M1175, also leaving lateral portions of the nodulus intact.


View larger version (68K):
[in this window]
[in a new window]
  		FIG. 3. Scheme as in Fig. 2. Sagittal sections through the partial lesion in M9303. Central portions of the nodulus and rostral-ventral uvula within 2 mm of the midline were removed. A and B: at 0.5 mm, only lobule 9a of the uvula was intact (arrows). At 2.5 and 3.5 mm, lateral portions of nodulus and lobule 9d of the uvula are visible. Arrows point to the residual nodulus. C: summary of the reconstruction showing the intact nodulus in Zones 3 and 4 (arrows).


View larger version (20K):
[in this window]
[in a new window]
  		FIG. 14. A: maximal cross-coupled vertical (upward) velocities during yaw axis OKAN with animals tilted at various angles between 0° (upright) and 90° (on-side) in M1175, before (open circle ) and after (bullet ) the partial nodulo-uvulectomy shown in B (see Cohen et al. 1992, for a more complete description of lesion). Before lesion, peak vertical velocities were close to the velocity of stimulation (60°/s). After lesion, they fell to ~14°/s.

The complete and partial lesions are compared in the schematic of Fig. 4, ad related to the histochemically defined zones of Voogd et al. (1996). All of the nodulus and lobules 9c and 9d of the uvula were ablated in the complete lesions, whereas all or most of Zone 1 and most of Zone 2 of the nodulus and lobules 9c and 9d of the uvula were ablated in the partial lesions. This left a margin of intact nodulus and ventral uvula in Zone 2 and all of Zones 3 and 4 in the partial animals.

Complete nodulo-uvulectomy and the time constants of velocity storage

Time constants of the velocity storage component of the aVOR are dependent on orientation of the head relative to gravity in normal monkeys, and the horizontal and vertical/roll components vary inversely. In each case, the time constants are maximal when animals are positioned so that the axis of the angular stimulus is coincident with the spatial vertical (Angelaki and Hess 1994a,b, 1995; Dai et al. 1991, 1992; Matsuo and Cohen 1984; Raphan and Cohen 1986, 1988). Thus horizontal OKAN time constants are maximal in the upright position and decline as a function of tilt angle with animals side down, prone or supine (Dai et al. 1991; Wearne et al. 1996, 1997b). Postrotatory vestibular horizontal time constants are similarly reduced in side-down positions (Angelaki and Hess 1994a, 1995; Benson 1974; Guedry 1974; Raphan et al. 1992; Waespe et al. 1985). In contrast, vertical and roll OKAN and vestibular time constants are maximal in side-down or prone and supine positions (Dai et al. 1992), whereas the pitch and roll OKAN time constants become short when monkeys are upright (Matsuo and Cohen 1984; Raphan and Cohen 1988). This implies that the time constants of the velocity storage component of the pitch and roll aVOR are also short in the upright position.

After nodulo-uvulectomy in M9312, central horizontal velocity storage time constants, determined from OKAN and from per- and postrotatory nystagmus during rotation at60°/s with the animal in an upright position, increased on average from 12.0 s before to 23.6 s (Table 1; average of right and left values from Table 2). Horizontal aVOR time constants also increased on average from 29.3 to 59.2 s in the three other animals with complete nodulo-uvulectomy (Table 1). In contrast, vertical and roll time constants were short in every orientation after lesion (Fig. 5, B and D). On average, the aVOR vertical time constant for rotation about a spatially vertical axis fell in M9312 from 10.1 to 3.5 s postoperatively, and the roll time constant from 13.4 to 2.7 s (Table 2). Vertical aVOR time constants fell on average in M1107 and M1152 from 18.0 to 5.9 s (Table 1). These effects were enduring and were present on repeated testing >1 yr after lesion.

 
View this table:
[in this window] [in a new window]
  		TABLE 2. Complete nodulo-uvulectomy (M9312); OKAN and aVOR time constants about a vertical axis


View larger version (50K):
[in this window]
[in a new window]
  		FIG. 5. Vertical (A, B, E, and F) and roll (C and D) angular vestibuloocular reflex (aVOR) time constants before and after complete (A-D) and partial (E and F) nodulectomy. As shown in the insets, animals were rotated left ear down or prone about a spatial vertical axis. Top trace in each part is yaw axis position, recorded with a potentiometer that resent every 360°. This gives the "sawtooth" appearance during constant velocity rotation. The 2nd trace is desaccaded vertical (omega X) or roll (omega Y) eye velocity. The central (velocity storage) time constant is listed over each trace. It was obtained by a double exponential fit with the cupula time constant constrained to 4 s. Note the up-down asymmetry in the preoperative vertical eye velocity. Postoperatively, the time constants for upward and downward velocities were similarly short. The central time constant of roll after lesion fell to between 1 and 3 s.

The dependence of the OKAN time constants on gravity was also lost after nodulo-uvulectomy. Regardless of head position, the horizontal and vertical time constants remained close to those previously recorded in the upright position. The failure of reorientation of the head to affect the horizontal time constant after nodulo-uvulectomy has been demonstrated previously (Angelaki and Hess 1995; Waespe et al. 1985). The change in the sensitivity of the vertical time constant to gravity is shown in Fig. 6. Before lesion, M9312 had strong upward OKN and OKAN in the left ear down position (-omega X; Fig. 6A), with an OKAN time constant of 13.3 s. Postoperatively, upward OKN was equally vigorous, but the on-side time constant of upward OKAN was now 2.5 s (-omega X; Fig. 6B). This is close to the mean upward OKAN time constant recorded in four cynomolgus monkeys when sitting erect (3.5 s) (see Fig. 6A of Dai et al. 1991).


View larger version (54K):
[in this window]
[in a new window]
  		FIG. 6. Vertical optokinetic nystagmus (OKN) and OKAN before (A) and after (B) complete nodulo-uvulectomy. The animal was on its side, 90° right ear down (RED), and received upward optokinetic stimulation in pitch about a vertical axis (see inset). Preoperatively, there was strong upward OKAN, which decayed with a time constant of 13.3 s. Postoperatively, OKAN decayed rapidly with a time constant of 2.5 s.

In conjunction with the increase in the horizontal time constant, each of the completely lesioned animals developed periodic alternating nystagmus (PAN; Table 1) (Waespe et al. 1985). The PAN was somewhat asymmetric in M9312, being shorter to the left than to the right, with a mean period of 170 s (Fig. 7). The velocity vector of the nystagmus was predominantly horizontal (omega Z), but there was a small torsional component (omega Y) that tipped the vector back in the head, as well as nonperiodic spontaneous downward nystagmus. Horizontal PAN was also present in each of the other three animals with complete lesions (Table 1). It developed spontaneously if animals were in darkness, but was also elicited by low-amplitude vestibular or optokinetic stimuli. The periods ranged from 150 to 250 s. PAN was suppressed in light. It disappeared with parenteral administration of small doses of baclofen (0.5-1 mg/kg) (Halmagyi et al. 1980), a gamma -aminobutyric acid-B (GABAB) agonist that shortens the dominant horizontal time constant of velocity storage (Cohen et al. 1987). This is consistent with the postulate of Leigh et al. (1981) that the oscillation of PAN is produced by large increases in the time constant of brain stem integrators, unchecked by inhibition.


View larger version (57K):
[in this window]
[in a new window]
  		FIG. 7. Periodic alternating nystagms (PAN) following complete nodulo-uvulectomy in M9312. Nystagmus was initiated by a step of velocity of 60°/s in yaw around a yaw axis. Continuous oscillations developed in horizontal (omega Z) and roll (omega Y) eye velocity with a period of ~150 s. A sustained downward vertical nystagmus (omega X) was also present.

Completely lesioned animals also lost the ability to reduce or "dump" horizontal slow-phase velocity during per- or postrotatory nystagmus and OKAN when they were exposed to a self-stationary visual surround (Waespe et al. 1985) (Table 1). In the example shown in Fig. 8A, horizontal eye velocity (omega Z) rapidly dropped by ~70% during the visual exposure, presumably due to activation of direct visual-oculomotor pathways that suppressed the nystagmus. Eye velocity returned to close to preexposure levels after the animal was once again in darkness and decayed with the same long time constant as the response in darkness. A small loss of stored velocity in Fig. 8A is expected, due to visual feedback from the eyes moving over the stationary surround during the period of suppression, rather than to activation of the dump mechanism (Waespe et al. 1985). Loss of control of the horizontal time constant in tilt- and light-dumping paradigms after nodulo-uvulectomy was present in each of the completely lesioned animals (Table 1).


View larger version (47K):
[in this window]
[in a new window]
  		FIG. 8. Postoperative light-dumping in complete (A; M9312) and partially lesioned (B; M9303) animals during per-rotatory nystagmus. The animals and the surround were rotated at 60°/s in the same direction in darkness. The light was switched on (top trace, light on) for 4 s, exposing a self-stationary visual surround. There was little loss of stored eye velocity in the completely lesioned animal (A), whereas eye velocity had almost disappeared in the partially lesioned animal (B). Note the long horizontal time constant in the completely lesioned animal.

Complete nodulo-uvulectomy and spatial orientation of the aVOR

During eccentric rotation (centrifugation) in the normal animal, the GIA vector tilts relative to the upright head. This has two consequences: the horizontal velocity storage time constant is reduced, and vertical cross-coupled components of eye velocity appear, tending to align the eye velocity axis with the GIA (Raphan et al. 1996; Wearne et al. 1996, 1997b). The completely lesioned animal lost both of these features. In the example shown in Fig. 9, the horizontal time constant for centered rotation was 13.5 s (Fig. 9B). Horizontal time constants for facing and back to motion centrifugation, which tilted the GIA by 51° about the nasooccipital axis, were the same whether facing motion (13.1 s, Fig. 9A) or back to motion (12.6 s, Fig. 9C). In addition, spatial orientation was lost, and there was no vertical (omega X) or torsional (omega Y) cross-coupling (Fig. 9, A and C). A small downward vertical component, present to the same extent during centered rotation (Fig. 9B, arrow A) as during facing and back to motion centrifugation (B and C), was probably spontaneous nystagmus. Ocular counter-rolling in response to the tilt of the GIA about the nasooccipital axis was preserved (Fig. 9, A and C; arrows B, psi ). It should be noted that the horizontal time constant was ~13 s in response to rotation at 400°/s during centered and off-center rotation (Fig. 9). This was substantially less than the average aVOR time constants of this animal when rotated at 60°/s (Table 2). The difference is due to the reduction in horizontal time constant as rotational velocity increases. This relationship originally described in normal animals (Raphan et al. 1979), was present after both partial (Cohen et al. 1992) and complete nodulo-uvulectomy. Thus it does not depend on the integrity of the nodulus and uvula.


View larger version (38K):
[in this window]
[in a new window]
  		FIG. 9. Postoperative responses of M9312 to -Z yaw axis centrifugation. Eye position and eye velocity responses during centrifugation (A and C) and centered rotation (B) in yaw with an angular acceleration of 40°/s2 up to a peak angular velocity of 400°/s. Three-dimensional eye position is represented using Euler angles in the Fick rotation convention, with (phi , theta , psi ) designating rotations about the yaw axis, the rotated pitch axis, and the optic (roll) axis in that order. Pitch (omega X), roll (omega Y), and yaw (omega Z) components of the eye velocity vector are displayed in head coordinates. Top panel shows the angular and linear stimulus profiles over time: the angular velocity of the centrifuge (ANG VEL) is indicated with a light line and ranges from 0 to 400°/s; the corresponding tilt angle of the GIA is indicated with a heavier line. For centered rotation, the GIA tilt remains at 0°. A-C: time constant of horizontal eye velocity (omega Z) was the same when facing or back to motion (A and C) as in centered rotation (B). Therefore it was unaffected by the tilted GIA. No cross-coupling from horizontal to vertical was present (compare arrows A in A and C with arrow A in B). During centrifugation, torsional eye position (psi ), shifted in the direction of theGIA tilt, indicating normal counter-rolling (arrows B).

Spatial orientation in the pitch plane was tested by radial centrifugation (facing nose in or out, Fig. 10). Before operation, a prominent cross-coupled roll component appeared during rotation, building to a peak after the angular acceleration had ended (omega Y; Fig. 10A, arrowheads). In accordance with the backward tilt of the GIA, the roll component reversed during clockwise and counterclockwise rotation in the nose-out orientation. The time constant of the roll component was long and was approximately equal to the time constant of the horizontal component (omega Z). A small vertical component (omega X) was also present during acceleration. The peak horizontal eye velocity was 140°/s, less than would be expected given the acceleration and peak velocity of the stimulus. This lower value was likely due to the reduction in the horizontal time constant produced by the tilted GIA.After lesion, the peak horizontal eye velocity reached 200°/s.This higher peak velocity was consistent with the loss of sensitivity of the horizontal time constant to the tilted GIA (omega Z; Fig. 10C). There was a transient increase in roll eye velocity during the angular acceleration, concurrent with activation of the semicircular canals, but it decayed to zero when constant angular velocity was reached; in the constant angular velocity phase, there was no cross-coupling to roll.


View larger version (44K):
[in this window]
[in a new window]
  		FIG. 10. Pre- (A) and postoperative (B) centrifugation in M9312 in the radial orientation with the nose out. A: horizontal, vertical, and roll eye velocities for both directions of rotation. Centrifugation induced a positive cross-coupled roll component during -Z rotation and a negative roll component during +Z rotation (arrowheads). B: eye velocity axis plotted in 3 dimensions (open circle ) with superimposed fitted eigenvector (heavy black line). Each graph is composed of 2 trials, one for +Z and another for -Z rotation. Circles start at the points farthest from the origin on the top and bottom halves of each graph and proceed toward zero. The eye rotation axis and the eigenvector both tilted back in the head by ~35-40°, slightly underestimating the tilt of the GIA (indicated on inset showing monkey's head). The components of the fitted eigenvectors for +Z and -Z horizontal eye velocity are shown in the inset to the far right of the 3-D graph. C: after nodulo-uvulectomy, eye velocity reached a higher peak due to the longer horizontal time constant. A roll component developed during angular acceleration in the direction opposite to that required to align the eye velocity with the GIA. This component decayed quickly and did not affect the final eye velocity trajectory. D: in 3 dimensions, the fitted eigenvector remained close to alignment with the Z-axis.

The loss of spatial orientation is illustrated in the three-dimensional plots of the eye velocity trajectory representing a complete data set in this animal (Fig. 10, B and D). Preoperatively, the eye velocity axis tilted back in the head, aligning closely with the GIA for both directions of rotation. The fitted eigenvector (Fig. 10B, heavy line) is shown superimposed on the three-dimensional eye velocity trajectory (open circle ). The components of the fitted eigenvectors for the +omega Z and -omega Z eye velocities to the right of the three-dimensional graph demonstrate the alignment of the eigenvector with the GIA. It tilted back in the head by 38° for +omega Z eye velocities and 36° for -omega Z eye velocities. This underestimated the 51° backward tilt of the GIA. Postoperatively, the eye velocity axis remained aligned with the Z-axis of the head (Fig. 10D, open circle ). The fitted eigenvector (heavy line) reflected this alignment. The X and Y components of the eigenvector were close to zero for both directions of rotation (+omega Z: X, -0.06; Y, 0.07; -omega Z: X, 0.02; Y, 0.01) and the Z component was close to 1 (0.99 and -0.99), producing negligible tilt of the eigenvector.

Loss of axis reorientation for GIA tilts elicited in each cardinal head plane is shown in two-dimensional planar projections of eye velocity for tangential (Fig. 11A) and radial centrifugation (Fig. 11B). Before lesion, facing and back to motion (tangential) centrifugation induced roll-plane tilts of the eye velocity axis in the same direction as the GIA tilt (Fig. 11A, black-square). Axis shifts in the roll plane were asymmetric, producing larger shifts during upward (-X) than downward (+X) cross-coupling. With right ear out, the axis tilted by 50 and 45° for upward and downward cross-coupling. With left ear out, it tilted by 32 and 25° for upward and downward coupling. Nose out in the radial orientation, the eye velocity axis tilted by 38° for +omega Z and by 36° for -omega Z eye velocities, and when nose in, by 70° for +omega Z and by 34° for -omega Z eye velocities (Fig. 10, bottom row, black-square). After lesion (Fig. 11, open circle ), the eye velocity axis remained aligned with the head Z-axis during eccentric rotation, irrespective of the direction of GIA tilt, and there were only negligible tilts of the fitted eigenvector in both roll (A) and pitch (B) planes, similar to the alignment during centered rotation.

Thus complete nodulo-uvulectomy abolished axis reorientation of velocity storage associated with tilts of the GIA in any direction. This was true for all paradigms. Axis reorientation was not specifically tested with centrifugation in the other animals with complete nodulo-uvulectomy (M1107, M1152, and M1153), nor was ocular torsion measured. As noted, however, these animals had longer horizontal time constants and reduced vertical time constants after lesion, and their yaw axis time constants did not change when the animals were placed in positions that tilted the GIA with regard to the head during horizontal postrotatory nystagmus or OKAN (Waespe et al. 1985). This corroborates the loss of spatial orientation after complete nodulo-uvulectomy in these animals.

Partial nodulo-uvulectomy and time constants of velocity storage

After partial lesions that involved the central portions of the nodulus and uvula, animals retained control of the horizontal time constants, but lost control of the vertical and roll time constants. Thus the pitch and roll aVOR time constants were reduced for stimulation about a spatial vertical axis, as in the completely lesioned animals. Examples of reduction in the pitch aVOR time constant before and 1 wk after surgery are shown in Fig. 5, E and F; there was no recovery 4 mo later. The central (velocity storage) time constants of the vertical aVOR, recorded on-side in M9303, were reduced from 10.9 s to 1.7 ± 0.6 (SD) s for upward and from 8.6 s to 2.6 ± 0.4 s for downward slow phase velocities (Table 3). During velocity steps of 60°/s, either nose down (prone) or nose up (supine), the time constant of the roll VOR for both directions of rotation was reduced from an average 7.4 s before surgery to 2.2 s after surgery (Table 3).

 
View this table:
[in this window] [in a new window]
  		TABLE 3. Partial nodulo-uvulectomy (M9303); aVOR time constants about a vertical axis

In contrast, horizontal time constants were the same before and after partial lesions. Horizontal OKAN time constants were 14 s on average for leftward and rightward eye velocities in M9303 before operation versus 15.5 s after operation (Fig. 12, Tilt angle 0°). The horizontal aVOR time constants were 13.7 s in the first preoperative test and 12.5 s in the postoperative tests in this animal (Table 3). Similarly, the horizontal aVOR time constants were 26.5 s on average in the other partially lesioned animals before operation (M1173 and M1175) and 24.0 s after operation (Table 1). None of the partially lesioned animals had periodic alternating nystagmus after surgery.


View larger version (21K):
[in this window]
[in a new window]
  		FIG. 12. Pre- and postoperative horizontal OKAN time constants in M9303 as a function of head tilts, 0, 45, and 90° to LED and RED. Slope parameters were obtained from 1st-order linear regressions. Each data point is the average of 2 measurements. A: rightward OKAN eye velocities. Preoperatively (open circle ), the time constant was ~14 s in the upright position, and fell with a slope of -0.09 s/deg. Postoperatively (black-triangle), the upright time constant was several seconds longer, but the slope was not reduced by the partial lesion (-0.1 s/deg). B: leftward OKAN eye velocities. Preoperatively (black-triangle), the slope was -0.6 s/deg. Postoperatively (open circle ), the upright time constant was similar to preoperative values, whereas the slope (-0.03) was slightly reduced from the preoperative value.

The ability to reduce the horizontal time constant when the GIA was tilted with respect to the head was quantified in M9303 by measuring the horizontal OKAN time constant with the head tilted at angles of 0, 45, and 90° (Fig. 12). Before operation, the time constant for eye velocity to the right fell from 14.5 s at 0° to 9 s at 90°, with a slope of -0.09 s/deg of tilt (Fig. 12A). Postoperatively, the upright time constant was slightly longer (2.5 s), but the slope was the same (-0.1 s/deg vs. 0.09 s/deg). For leftward eye velocities (Fig. 12B), the preoperative slope was -0.06s/deg. Postoperatively, the slope was slightly less (-0.03s/deg). The ability to discharge velocity storage in response to rapid reorientation of the head during postrotatory nystagmus was also present in the other partially lesioned monkeys (Table 1).

All three animals with partial nodulo-uvulectomy were also able to discharge slow-phase eye velocity, i.e., to dump activity responsible for horizontal slow-phase eye velocity during conflicting visual stimuli (Table 1). Exposure to a self-stationary visual surround during steps of angular velocity in M9303 produced a rapid drop in eye velocity and a loss of slow-phase velocity when the animal was once again back in darkness (Fig. 8B). In this instance, the 3.5-s period of suppression had resulted in an almost complete loss or dump of velocity storage. When the period of exposure was extended to 5 s, there was no nystagmus when the animal was once again in darkness. The ability of the partially lesioned animals to dump stored eye velocity as a result of exposure to a self-stationary visual surround is in contras