INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The vestibular system is classically associated with detecting the motion of the head-in-space to generate the reflexes that are crucial for our daily activities, such as stabilizing gaze (gaze = eye-in-head + head-in-space) via the vestibuloocular reflex (VOR) during walking and running (Grossman et al. 1988, 1989). During passive whole-body rotation in head-restrained animals, vestibular afferents originating in the semicircular canals encode the angular velocity of the head-in-space (Goldberg and Fernandez 1971). These vestibular afferents, in turn, project to second-order neurons within the vestibular nuclei, which encode angular head velocity during the compensatory eye movements generated by the VOR (Cullen and McCrea 1993; McCrea et al. 1987; Scudder and Fuchs 1992). However, in addition to input from the vestibular nerve, the vestibular nuclei receive projections from many structures that could influence their discharge. For example, neurons within oculomotor/gaze control pathways, such as the saccadic burst neurons in the paramedian pontine reticular formation, send direct projections to the vestibular nuclei (Sasaki and Shimazu 1981). Furthermore, neck muscle spindle afferents are known to influence the activity of vestibular nuclei neurons in decerebrate animals (Anastasopoulos and Mergner 1982; Boyle and Pompeiano 1981; Fuller 1988; Wilson et al. 1990) via a disynaptic pathway (Sato et al. 1997). Moreover, cortical areas, which have been implicated in more cognitive aspects of vestibular function including the perception of spatial orientation, the ability to navigate (in the absence of visual cues), and gaze control, project to the vestibular nuclei (for review, see Fukushima 1997). Thus given the convergence of multiple inputs to the vestibular nuclei, it is natural to ask how the signals carried by these inputs are integrated during our daily activities.

Here we have focused on the behavior of a distinct population of vestibular nuclei neurons termed position-vestibular-pause (PVP) neurons during horizontal head rotations. Type I PVP neurons are thought to constitute most of the intermediate leg of the direct VOR pathway; they receive a strong monosynaptic connection from the ipsilateral semicircular canal afferents and, in turn, project directly to the extraocular motoneurons (Cullen and McCrea 1993; McCrea et al. 1987; Scudder and Fuchs 1992). These neurons derive their name from the signals they carry during head-restrained oculomotor and vestibular paradigms: their firing rates increase when the eyes move to more contralaterally directed positions; during slow phase vestibular nystagmus, these neurons are sensitive to ipsilateral head rotations (i.e., a type I response); and their discharges cease (pause) for ipsilaterally directed saccades and vestibular quick phases. Type II PVP neurons have oppositely directed eye- and head-motion sensitivities to those of type I PVP neurons, and their role in the processing of vestibular information is less well understood. Nevertheless, in general, they behave in a quantitatively similar manner to type I PVP neurons during head-restrained rotations and eye movements (Scudder and Fuchs 1992).

The first goal of this study was to address the general question of whether the PVP neuron responses to active and/or passive movements are modified in a manner that depends on the animal's current behavioral goal. For example, the response of PVP neurons might be selectively attenuated for gaze redirection versus stabilization, consistent with their role in mediating the VOR. Alternatively, it is also possible that PVP neurons might differentially encode head velocity during self-generated versus passively applied head-in-space rotations and/or during rotation of the head and body together in space versus rotation of the head relative to an earth-stationary body.

On the one hand, there is already much evidence to suggest that type I PVP neurons differentially encode head-velocity during gaze redirection versus gaze stabilization. First, while type I PVP neurons encode head velocity during the compensatory slow phase component of the VOR evoked by passive whole-body rotation, they pause or cease firing during vestibular quick phases where gaze is redirected (Cullen and McCrea 1993; Fuchs and Kimm 1975; Keller and Daniels 1975; Keller and Kamath 1975; Lisberger et al. 1994a,b; McConville et al. 1996; McCrea et al. 1987; Miles 1975; Roy and Cullen 1998; Scudder and Fuchs 1992; Tomlinson and Robinson 1984). Second, when gaze is voluntarily redirected using coordinated eye-head gaze shifts, the head-velocity signal carried by type I PVP neurons is significantly attenuated as compared with passive whole-body rotation (Cullen and McCrea 1993; McCrea et al. 1996; Roy and Cullen 1998) in an amplitude-dependent manner (Roy and Cullen 1998). Third, type I PVP neuron responses are attenuated by ~30% as compared with passive rotation in the dark when monkeys suppress their VOR (and therefore redirect their gaze to move with the head in space) during passive whole-body rotation by tracking a target that moves with the head (Cullen and McCrea 1993; McCrea et al. 1996; Roy and Cullen 1998; Scudder and Fuchs 1992).

On the other hand, whether PVP neurons differentially encode head-velocity during self-generated versus passively applied rotations of the head-in-space is less clear. It has been proposed that an enhanced sensitivity of the VOR pathways during active head movements might increase VOR responses in humans as compared with passive head motion when the goal is to stabilize gaze (Demer et al. 1993; Jell et al. 1988). However, this behavioral observation remains to be confirmed at the neural level. We have previously shown (Roy and Cullen 1998) that type I PVP neuron discharges in the rhesus monkey are similar during the VOR elicited by passive whole-body rotation and during the active head movements made in the time interval that immediately follows a gaze shift. In this interval, an ocular counter roll compensates for the residual active head motion such that the monkey's axis of gaze is stable relative to space. Gdowski and McCrea (1999) also reported that the majority of PVP neurons in squirrel monkeys encode head-in-space motion during simultaneous active and passive head motion. However, they emphasized that 35% of the neurons were better related to the passive component of the motion than to the total head-in-space motion during this same paradigm. While this latter result is actually the opposite of what would be expected if the efficacy of the VOR was, in fact, enhanced during active head movements, its interpretation is limited given that the gaze goal of the monkeys was not reported.

The second goal of the present study was to determine the neural mechanism(s) that contribute to the differential processing of head-velocity information by PVP neurons. For example, during gaze shifts, there are several possible mechanisms that could modulate PVP neuron discharges. It is likely that inputs from the brain stem saccadic burst generator to the vestibular nuclei could function to suppress PVP neuron responses during voluntary combined eye-head gaze shifts (Roy and Cullen 1998). Specifically, burst neurons in the paramedian pontine reticular formation project to type II neurons in the vestibular nucleus (Sasaki and Shimazu 1981), which in turn send an inhibitory projection to type I PVP neurons (Nakao et al. 1982). This pathway almost certainly provides a powerful inhibitory drive to type I PVP neurons during rapid gaze shifts. It is also possible that inhibitory inputs from neck proprioceptors contribute to the attenuation of PVP neuron discharges. Studies in anesthetized animals have shown that the activation of neck proprioceptors can influence the activity of vestibular nuclei neurons (Anastasopoulos and Mergner 1982; Boyle and Pompeiano 1981; Fuller 1988; Wilson et al. 1990). Furthermore, McCrea and colleagues recently reported that most if not all secondary vestibular neurons (including type I PVP neurons) in squirrel monkey are sensitive to passive neck rotation (Gdowski and McCrea 1999, 2000; McCrea et al. 1996). Finally, a signal related to the voluntary head motion itself, such as an efferent copy of the motor command to the neck musculature (McCrea et al. 1996) or a cortically derived signal representing the monkey's self-generated head motion, could influence the responses of PVP neurons.

To determine how PVP neurons process head-velocity information across a wide variety of behaviors and to understand the mechanisms that underlie the observed differential processing, we devised a sequence of paradigms in which the gaze goal was systematically varied for externally applied and/or self-generated head movements. We first characterized the discharges of PVP neurons in the head-restrained condition during passive whole-body rotation when gaze was stable (VOR) and when gaze was redirected (VOR cancellation paradigm). The neuronal discharges were then recorded during different gaze control tasks while the monkey experienced passive rotations of its head-on-body, generated voluntary head-on-body movements to orient to novel targets or track a slowly moving target, was passively rotated while simultaneously generating active head movements, and voluntarily "drove" its head and body together relative to space. We found that, during active and/or passive head movements, type I PVP neurons robustly encoded head velocity whenever monkeys stabilized their gaze relative to space, and were similarly attenuated during gaze-redirection tasks. Furthermore, the responses of type II PVP neurons were quantitatively comparable to those of type I PVP neurons during most behavioral conditions. Our results support the hypothesis that an efference copy of the brain stem oculomotor/gaze commands to redirect the visual axis in space underlies the "on-line" reduction in VOR pathway modulation when the VOR is functionally inappropriate.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Three rhesus monkeys (Macaca mulatta) were prepared for chronic extracellular recording using aseptic surgical techniques. All experimental protocols were approved by the McGill University Animal Care Committee and were in compliance with the guidelines of the Canadian Council on Animal Care.

Surgical procedures

The surgical techniques were similar to those previously described by Roy and Cullen (2001). Briefly, an 18- to 19-mm-diam eye coil (3 loops of Teflon-coated stainless steel wire) was implanted in the right eye behind the conjunctiva. In addition, a dental acrylic implant was fastened to each animal's skull using stainless steel screws. The implant held in place a stainless steel post used to restrain the animal's head, and a stainless steel recording chamber that was positioned to access the medial vestibular nucleus (posterior and lateral angles of 30°). During the surgery isoflurane gas was utilized to initiate (2-3%) and maintain (0.8-1.5%) anesthesia. After the surgery, buprenorphine (0.01 mg/kg im) was utilized for postoperative analgesia, and monkeys were allowed to recover for 2 wk before commencing experimental sessions.

Data acquisition

At the onset of each experiment, the monkey sat comfortably in a primate chair, which was placed on a vestibular turntable. With the monkey initially head-restrained, extracellular single-unit activity was recorded using enamel-insulated tungsten microelectrodes (7-10 M impedance, Frederick-Haer) as has been described elsewhere (Roy and Cullen 2001). The abducens nucleus, which was identified based on its stereotypical discharge patterns during eye movements (Cullen et al. 1993; Sylvestre and Cullen 1999), was located and used as a landmark to determine the location of the medial and lateral vestibular nuclei. Gaze and head position were measured using the magnetic search coil technique (Fuchs and Robinson 1966), and turntable velocity was measured using an angular velocity sensor (Watson). Unit activity, horizontal and vertical gaze and head positions, target position, and table velocity were recorded on DAT tape for later playback. Action potentials were discriminated during playback using a windowing circuit (BAK) that was manually set to generate a pulse coincident with the rising phase of each action potential. Gaze position, head position, target position, and table velocity signals were low-pass filtered at 250 Hz (8 pole Bessel filter) and sampled at 1,000 Hz.

Behavioral paradigms

Using juice as a reward, monkeys were trained to follow a target light (HeNe laser) that was projected, via a system of two galvanometer controlled mirrors, onto a cylindrical screen located 60 cm away from the monkey's head. Eye-motion sensitivities to saccades and ocular fixation were characterized by having the head-restrained monkey attend to a target that stepped between horizontal positions over a range of ±30°. To determine neuronal eye-motion sensitivities during smooth pursuit, head-restrained monkeys tracked sinusoidal (0.5 Hz, 80°/s peak velocity) target motion in the horizontal plane. Head-velocity sensitivities to passive whole body rotation (0.5 Hz, 80°/s peak velocity) were tested by rotating monkeys about an earth vertical axis in the dark [passive whole-body rotation (pWBR)] and while they cancelled their VOR by fixating a target that moved with the vestibular turntable (pWBRc). Target and turntable motion, and on-line data displays were controlled by a UNIX-based real-time data-acquisition system (REX) (Hayes et al. 1982).

After a neuron was fully characterized in the head-restrained condition, the monkey's head was slowly and carefully released to maintain isolation. Once released, the monkey was able to rotate its head through the natural range of motion in the yaw (horizontal), pitch (vertical), and roll (torsional) axes. The response of the same neuron was then recorded during the voluntary head movements made during combined eye-head gaze shifts (15-65° in amplitude) and during combined eye-head gaze pursuit of a sinusoidal target (0.5 Hz, 80°/s peak velocity). In addition, neuronal responses to combined passive and active head motion were recorded while head-unrestrained monkeys were passively rotated (0.5 Hz., 80°/s peak) and allowed to simultaneously generate voluntary head-on-body movements. A subset of neurons was tested during a "driving paradigm" in which the monkey moved its head and body together in space. During this paradigm, head-restrained monkeys manually manipulated a steering wheel to control the initiation of the movement as well as the rotational velocity of the turntable on which they were seated. The goal of the monkey was to align a chair mounted target with a moving laser target.

Finally, the influences of dynamic and static neck proprioceptive inputs on neural discharges were investigated. Two different paradigms were used to dynamically activate the neck afferents. First, the experimenter manually rotated the monkey's head to induce rapid motion of the head relative to a stationary body. Second, the monkey's head was held stationary relative to the earth while its body was passively rotated at 0.1, 0.2, 0.5, 1, 1.5, and 2 Hz at 20°/s peak velocity and 0.2, 0.5, 1, 1.5, and 2 Hz at 40°/s peak velocity. The gain of the cervicoocular reflex induced during the rotations was calculated as the resultant desaccaded eye velocity divided by the turntable velocity. To test for the influence of static neck afferent activation, the monkey's body was held at different static positions relative to its earth-stationary head, and the mean firing rate was calculated. During this testing, the torque produced by the monkey against the head-restraint was measured using a reaction torque transducer (Sensotec).

Analysis of neuron discharges

Before analysis, recorded gaze and head-position signals were digitally filtered at 125 Hz. Eye position was calculated from the difference between gaze and head-position signals. Gaze, eye, and head-position signals were digitally differentiated to produce velocity signals. The neural discharge was represented using a spike density function in which a Gaussian function was convolved with the spike train (SD of 5 ms for saccades and gaze shifts and 10 ms for remainder of the paradigms) (Cullen et al. 1996). Saccade and gaze shift onsets and offsets were defined using a ±20°/s gaze velocity criterion. Subsequent analysis was performed using custom algorithms (Matlab, Mathworks).

To quantify a neuron's response to eye movement, we analyzed periods of steady fixation to obtain a resting discharge (bias, sp/s) and an eye-position sensitivity [k_x, (sp/s)/°] and periods of saccade-free smooth pursuit to obtain a resting discharge (bias, sp/s), an eye-position sensitivity [k_sp, (sp/s)/°], and an eye-velocity sensitivity [r_sp, (sp/s)/(°/s)] using a multiple regression analysis (Roy and Cullen 1998). Spike trains were assessed to determine whether neurons paused or burst during saccades. In cases where neurons did burst, the resting discharge (bias, sp/s), eye position [k_sac, (sp/s)/°] and eye velocity [r_sac, (sp/s)/(°/s)] sensitivities were also estimated during saccades.

A least-squared regression analysis was then used to determine each neuron's phase shift relative to head velocity, resting discharge (bias, sp/s), and head velocity [g_pWBR (sp/s)/°/s] during pWBR and pWBRc. Only unit data from periods of slow-phase vestibular nystagmus during pWBR or steady fixation during pWBRc that occurred between quick phases of vestibular nystagmus and/or saccades were included in the analysis. A least-squared regression analysis was applied to neuronal discharges during active head-on-body motion, active head and body motion (driving paradigm), combined passive and active head motion, passive head-on-body rotations, and passive body-under-head rotations. The models utilized for each condition are described in RESULTS. To quantify the ability of the linear regression analysis to model neuronal discharges, the variance-accounted-for (VAF) provided by each regression equation was determined. The VAF was computed as {1 - [var(est  fr)/var(fr)]}, where est represents the modeled firing rate (i.e., regression equation estimate) and fr represents the actual firing rate. Note that only data for which the firing rate was greater than zero were included in the optimization. Statistical significance was determined using paired Student's t-tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The firing behaviors of two distinct classes of vestibular nuclei neurons are presented in the following text. First, we describe the responses of type I PVP neurons during a sequence of paradigms in which the gaze goal was systematically varied and the head movements were externally applied and/or self-generated. We then describe the responses of type II PVP neurons whose head and eye-velocity sensitivities in the head-restrained condition were opposite to those of type I PVP neurons, during each of the same conditions.

Type I PVP neurons

HEAD-RESTRAINED CHARACTERIZATION. The type I PVP neuron illustrated in Fig. 1 is typical of our sample (n = 24) in that its firing rate increased for contralaterally directed eye positions during spontaneous eye movements (Fig. 1A, see inset), and its firing rate phase lagged contralaterally directed eye velocity and led contralaterally directed eye position during smooth pursuit (Fig. 1B). During pWBR (Fig. 1C) the neuron increased its firing rate in response to ipsilateral head motion (i.e., a type I response). In addition, each type I PVP neuron stopped firing or "paused" during ipsilaterally directed saccades and vestibular nystagmus quick phases (vertical arrows in Fig. 1, A and C). Thus the type I PVP neurons in our sample were comparable to those that have been described in previous reports (Cullen and McCrea 1993; Fuchs and Kimm 1975; Keller and Daniels 1975; Keller and Kamath 1975; Roy and Cullen 1998; Scudder and Fuchs 1992). We also utilized a second pWBR paradigm in which the monkey cancelled its VOR by fixating a head-centered visual target that moved with the vestibular turntable (pWBRc; Fig. 1D). This VOR cancellation paradigm has been used extensively to dissociate a neuron's vestibular sensitivity from its eye-movement related modulation. As can be seen in Fig. 1D, type I PVP neurons remained well modulated in response to ipsilateral head velocity during this paradigm.

View larger version (24K): [in this window] [in a new window]   Fig. 1. Activity of an example type I position-vestibular-pause (PVP) neuron (unit b39_1) during the head-restrained condition. A: the neuron increased its discharge for contralaterally directed eye movements and paused for ipsilaterally directed saccades (vertical arrows). Inset: mean neuronal firing rate was well correlated with horizontal eye position during periods of steady fixation. B: the neuron also increased its discharge during contralaterally directed smooth pursuit. A model based on the bias discharge, the eye-position sensitivity, and the eye-velocity sensitivity of the neuron provided a good fit to the neural activity (smooth pursuit model, thick trace). C: passive whole-body rotation (pWBR) was used to characterize the neuron's response to head movements during VOR in the dark. A model based on the bias discharge, the eye-position sensitivity, and the head-velocity sensitivity during the compensatory eye movements made during pWBR (solid trace) is superimposed on a model fit that also included a head-acceleration term (gray shaded trace). The vertical arrows indicate pauses in activity for ipsilaterally directed vestibular quick phases. D: unit b39_1 was typical in that its modulation was less during pWBR while the monkey cancelled its VOR (pWBRc) by fixating a target that moved with the table (pWBRc estimate, thin trace) than during pWBR (pWBR model, thick trace). Traces directed upward are in the ipsilateral direction. E, eye position; H, head position; , eye-in-head velocity; , head-in-space velocity; , gaze velocity (=  + ); FR, firing rate.

VOLUNTARY HEAD-ON-BODY MOTION. Rapid gaze redirection. After the head-restrained characterization had been completed, the monkey's head was slowly released to allow a wide range of motion in all three axes (yaw, pitch, and roll). Unit activity was carefully monitored during the transition from the head-restrained to head-unrestrained condition to ensure that neurons remained isolated and undamaged.

View larger version (51K): [in this window] [in a new window]   Fig. 2. Activity of example type I PVP neurons (units b39_1 and cr81_1) during and after voluntary ipsilaterally directed combined eye-head gaze shifts. A: the pWBR model overpredicted the discharge of this neuron during small- and large-amplitude gaze shifts (pWBR model, thick trace, 2nd row from bottom). Estimate 1 provided a poor fit to the firing rate during gaze shifts (Estimate 1, thin trace, 2nd row from bottom). Estimate 2, with its additional eye-velocity term, provided a much improved fit to the neural activity (Estimate 2, thick trace, bottom row). Dotted vertical lines indicate the onset and offset of gaze shifts using a ±20°/s criterion. Open arrows indicate the post-gaze-shift intervals. B: during gaze shifts, the head-velocity sensitivity of our sample of type I PVP neurons, obtained using Estimate 2, decreased significantly as gaze shift amplitude increased from 15 to 65° (open columns), and for all gaze shift amplitudes, responses were significantly smaller than those resulting from pWBR (solid column). C: in contrast, head-velocity sensitivities in the post-gaze-shift interval (denoted by open arrows in A; gray shaded columns) were comparable to those measured during pWBR (solid column). Error bars show SE. Asterisks, P < 0.05. G, gaze position.

View larger version (28K): [in this window] [in a new window]   Fig. 3. Activity of an example type I PVP neuron (unit b39_1) during voluntary contralaterally directed head-restrained saccades and combined eye-head gaze shifts. A: during small (left) and large (right) saccades, the majority of type I PVP neurons did not pause or burst in activity. B: during small-amplitude gaze shifts (<35°, left), the pWBR model (thick trace) provided a good prediction of neural discharge. Similarly, during large-amplitude gaze shifts (right), the model provided a good fit until the discharge was driven to 0 by the faster head velocities.

View larger version (27K): [in this window] [in a new window]   Fig. 4. Responses of a typical type I PVP neuron (unit b39_1) to voluntary head-on-body motion during combined eye-head gaze pursuit. A: the neuron's response to self-generated head motion was overpredicted by the neuron's sensitivity during pWBR (pWBR model, thick trace, 2nd row from bottom). Estimate 2 provided a good of the firing rate (Estimate 2, thick trace, bottom row) and the estimated head-velocity sensitivity was significantly reduced as compared with pWBR. B: on a neuron-by-neuron basis, the eye-velocity sensitivities estimated during gaze pursuit were similar to those estimated during smooth pursuit. Dotted line represents unity (slope = 1). : target velocity.

Do type I PVP neurons differentially encode head velocity during self-generated vs. passively applied rotations of the head-in-space?

Both gaze shifts and gaze pursuit involve the voluntary movement of the head on the body. To determine whether type I PVP neurons might differentially encode head-velocity during self-generated versus passively applied rotations of the head-in-space, we used a paradigm in which head-restrained monkeys voluntarily drove or controlled the direction and rotation velocity of the turntable, thus moving both their heads and bodies together in space. The monkeys were trained to align a turntable mounted laser target (T_table) with a computer controlled target (T_goal; Fig. 5A, see schema), which either stepped from one location to another or moved sinusoidally. The data traces shown in Fig. 5A illustrate the discharge activity of an example type I PVP neuron during the two behavioral tasks. When the monkey redirected its gaze ipsilaterally to align the turntable with a target that stepped (Fig. 5A; middle, ), the pWBR model over-predicted the discharge of the neuron (pWBR model, thick trace). Indeed, for the neurons tested (n = 8), the estimated head-velocity sensitivity was significantly attenuated relative to pWBR to a level comparable to that observed during gaze shifts [Estimate 2 mean normalized g_est = 0.18 ± 0.21 (sp/s)/(°/s); Fig. 5B, ]. Similarly, when the monkey pursued the target (Fig. 5A; right), the pWBR model overpredicted the discharge of the neuron (pWBR model, thick trace), and the estimated head-velocity sensitivity was significantly reduced relative to pWBR [Estimate 2 mean normalized g_est = 0.64 ± 0.04 (sp/s)/(°/s); Fig. 5B, ]. This attenuation was comparable to that described for gaze pursuit and pWBRc above (P > 0.7 and > 0.27, respectively).

View larger version (41K): [in this window] [in a new window]   Fig. 5. Type I PVP neuron responses to voluntary combined head-body motion. A: head-restrained monkeys manually controlled a steering wheel to rotate the vestibular turntable relative to space (schema). Their goal was to align a turntable-fixed laser target (Ttable; schema) with a computer-controlled target (Tgoal; schema) that either stepped in location (middle) or moved sinusoidally (right panel). This example neuron (unit br23_1) was typical in that its response was well predicted by its sensitivity to head velocity during pWBR whenever the monkey stabilized gaze (pWBR model, thick trace). The neuron paused in activity for rapid ipsilaterally directed eye-head/body gaze redirections () and was less responsive to head motion when the target was pursued with both its eyes and head/body as compared with pWBR (right). B: type I PVP neurons had a greater attenuation during rapid gaze redirection () than during slow gaze redirection () and were not attenuated when gaze was stabilized () as compared with pWBR ().

In contrast, in the intervals where the monkey had acquired the new target, but the turntable was still moving (i.e., gaze was stable) the pWBR model provided a good fit to the firing rate (Fig. 5A, middle, pWBR model, thick trace; mean VAF = 0.64 ± 0.07) and accordingly the estimated head-velocity sensitivity was not significantly different from that during pWBR [mean normalized g_est = 0.92 ± 0.05 (sp/s)/(°/s); Fig. 5B, compare  and ]. This result is analogous to that described in the preceding text for the active head-on-body motion made when gaze is immobile immediately following gaze shifts. Thus the head-velocity information carried by type I PVP neurons was the same regardless of whether the head was voluntarily moved on the body or whether the head and body were voluntarily moved together; head-velocity sensitivities were similarly reduced when the monkey redirected its gaze and were unaltered when the monkey stabilized its gaze. Taken together, these findings support the hypothesis that type I PVP neuron responses to head motion vary in a manner that depends exclusively on the monkey's current gaze goal.

Type I PVP neurons

SIMULTANEOUS VOLUNTARY AND PASSIVE MOTION. In the behavioral tasks presented until this point, the head motion has been either passively imposed or self-generated, yet, under natural circumstances, passive perturbations of the head and/or body can occur at the same time as voluntary head motion. To further test the proposal that the head-velocity information carried by type I PVP neuron modulation depends only on current gaze goal, we characterized neuronal discharges during simultaneous passive and self-generated head motion. Head-unrestrained monkeys were encouraged to generate voluntary head-on-body movements (Fig. 6A, dashed line arrow in schema) while being passively whole-body rotated (Fig. 6A, solid arrow in schema). During this paradigm, the head-in-space motion is the sum of the passive turntable rotation and the active head-on-body motion.

View larger version (39K): [in this window] [in a new window]   Fig. 6. Type I PVP neuron responses to combined voluntary and passive head-in-space motion. A: head-unrestrained monkeys generated voluntary head-on-body movements (dashed arrow in schema) while being passively rotated by the vestibular turntable (thick arrow in schema). Head-in-space velocity (-in-space) is the sum of the passive rotation velocity (chair rotation) and voluntary head-on-body velocity (-on-body). The modulation of the example neuron (unit c131_1) was well correlated with the head-in-space motion (-in-space model, thick trace) whenever the monkey stabilized its gaze but was poorly related whenever the monkey rapidly redirected (vertical arrows) or slowly redirected its gaze (middle). B: the responses to both the passive and the voluntary components of head-in-space motion were significantly attenuated when the monkey either rapidly redirected (open columns) or slowly redirected (vertically striped columns) its gaze as compared with pWBR (solid column). In contrast, the neurons showed no attenuation during gaze stabilization (gray shaded columns).

INFLUENCE OF PASSIVE NECK PROPRIOCEPTOR ACTIVATION. To determine whether afferent inputs from neck muscle proprioceptors influence the activity of type I PVP neurons, we used two different paradigms. First we passively rotated the monkey's body while its head was held earth-stationary (Fig. 7A, see schema). The torque produced by the monkey against the head restraint was concurrently measured and found to be small (less than ±0.5 Nm) compared with that produced when orienting to food target (more than ±3.5 Nm). Thus during these passive rotations, the neck motor efference signals generated by the monkeys were minimal, yet the musculature was stretched such that neck proprioceptors were activated. The neuron shown in Fig. 7A was typical in that its activity was not significantly affected by the passive rotation of the neck. Its firing rate was well predicted by the pWBR model (note, in this case: head velocity = 0; pWBR model, thick trace). This finding is best appreciated when the firing rate is corrected for the neuron's eye-position sensitivity (Fig. 7A; FR_corr). For all neurons tested (n = 12), the neck-related signals were negligible [mean neck-velocity sensitivity = 0.07 ± 0.05 (sp/s)/(°/s)].

View larger version (31K): [in this window] [in a new window]   Fig. 7. Response of type I PVP neurons to passive neck rotation. A: the response of example neuron (unit gr11_2) was typical in that it was not modulated by neck-related signals when the monkey's body was passively rotated beneath its stationary head. The neural discharge was well described by the pWBR model (thick trace, 2nd row from bottom). This is emphasized when the firing rate and pWBR model are corrected for the neuron's eye-position sensitivity: FRcorr = FR - (kpWBR * E) and pWBR modelcorr = pWBR model  (kpWBR * E). B and C: during this task, the status of the cervicoocular reflex was simultaneously assessed by calculating the resulting gain (eye velocity/neck velocity). The gains of the monkeys tested are shown by the gray traces (triangle, monkey C; square, monkey G; diamond, monkey J) and were not significantly different from 0 over the range of frequencies and velocities (20°/s, B; 40°/s, C) tested. The average gain of the 3 monkeys is indicated by the solid black trace. The mean ± SE VOR gain (eye velocity/head velocity) is plotted for comparison (0.94 ± 0.06; see text). s, head-in-space velocity; s, body-in-space velocity; Torque: reaction torque produced by the monkey against the head-restraint.

View larger version (32K): [in this window] [in a new window]   Fig. 8. Response of type I PVP neurons to passive rotation of the head-on-body. A: the experimenter (hand in schema) passively rotated the monkey's head relative to its earth stationary body. The discharge of example neuron (unit gr11_2) was reliably predicted by the pWBR model whenever the monkey stabilized its gaze (thin trace). In contrast, the neuron paused in activity whenever the monkey rapidly redirected its gaze (). B: neuron responses during passive head-on-body rotation were comparable to those during pWBR when gaze was stable () and significantly attenuated when gaze was rapidly redirected () as compared with pWBR ().

View larger version (36K): [in this window] [in a new window]   Fig. 9. Response of type I PVP neurons to static body positions. The pWBR model accurately predicted the discharge activity of the example neuron (unit gr11_2) during stable gaze and body positions (thick trace, 2nd row from bottom). This is emphasized when the firing rate and pWBR model (bottom) are corrected for the eye-position sensitivity of the neuron (see Fig. 7 legend). Inset: the mean corrected firing rate was not correlated to static body position. Body: body position in space.

SUMMARY OF NEURAL DISCHARGE DURING PASSIVE AND VOLUNTARY HEAD MOTION. Type I PVP neurons responses to both passive and self-generated head motion were influenced by the gaze goal of the monkey. When the monkey stabilized its gaze, the head-velocity sensitivity of the neurons was comparable to that obtained during pWBR (Fig. 10; ). In contrast, when the monkey redirected its axis of gaze either slowly (Fig. 10; ) or rapidly (Fig. 10; ), the head-velocity sensitivity of the neurons was significantly attenuated as compared with pWBR. The neuronal responses were more attenuated during rapid gaze redirection than during slow gaze redirection (Fig. 10; compare  and ). Whether the head was passively or actively moved or whether the head moved relative to the body or not did not affect the neural discharges once the gaze goal was taken into account.

View larger version (28K): [in this window] [in a new window]   Fig. 10. Summary of type I PVP neuron discharge activity during passive and voluntary head motion. When the monkey redirected its axis of gaze either slowly () or rapidly (), the head-velocity sensitivity of the neurons was significantly attenuated as compared with pWBR (*, P < 0.05 relative to pWBR). The neuronal responses were significantly less during ipsilaterally directed rapid gaze redirection than during slow gaze redirection (, P < 0.05 relative to slow gaze redirection). In contrast, when the monkey stabilized its gaze, the head-velocity sensitivity of the neurons was comparable to that obtained during pWBR ().- - -, the average normalized head-velocity sensitivities across conditions with the same gaze goal.

Type II PVP neurons

In the present study, we also characterized type II PVP neurons. These neurons have opposite eye- and head-motion sensitivities to type I PVP neurons during the head-restrained paradigms shown in Fig. 1 and were included in the present report because they behaved very much like type I PVP neurons during each of the paradigms tested in our study. The only significant difference in firing behavior that we observed was that these neurons frequently discharged a burst during ipsilaterally directed saccades, vestibular quick phases, and gaze shifts. This difference is detailed in the following text.

The type II PVP neuron illustrated in Fig. 11 was typical of the neurons tested (n = 14) in that its firing rate increased for ipsilateral eye positions during spontaneous eye movements (Fig. 11A, see inset). The firing rate of type II PVP neurons was generally well correlated with eye position during ocular fixation (sample mean R² = 0.64 ± 0.06). The example neuron had a k_x of 1.31 (sp/s)/° [sample mean = 1.67 ± 0.37 (sp/s)/°] and a bias of 50 sp/s (sample mean = 55 ± 10 sp/s) during periods of fixation. During smooth pursuit, type II PVP neuron firing rate phase lagged ipsilateral eye velocity and led ipsilateral eye position (Fig. 11B). In general, neuron discharges were well described by the linear combination of eye velocity, eye position, and bias terms (sample mean VAF = 0.56 ± 0.09; Fig. 11B, smooth pursuit model, thick trace). The example neuron had a k_sp of 0.61 (sp/s)/° [sample mean = 1.52 ± 0.6 (sp/s)/°], a r_sp of 0.44 (sp/s)/(°/s) [sample mean = 0.54 ± 0.14 (sp/s)/(°/s)], and a bias_sp of 59 sp/s (sample mean = 64 ± 12 sp/s). The mean phase lag with respect to eye velocity for the sample of neurons during smooth pursuit was 64 ± 4.5°. During passive whole-body rotation in the dark (pWBR, Fig. 11C), the neuron increased its firing rate in response to contralateral head motion (i.e., a type II response). A model based on a combination of head velocity, eye position, and bias terms (Estimate 1) provided a good fit of type II PVP neuron activity during pWBR (sample mean VAF = 0.59 ± 0.06). The example neuron had a bias_pWBR of 67 sp/s (sample mean = 80 ± 10), a k_pWBR of 0.65 (sp/s)/° [sample mean = 1.25 ± 0.37 (sp/s)/°], and a g_pWBR of 1.71 (sp/s)/(°/s) [sample mean = 0.89 ± 0.08 (sp/s)/(°/s); Fig. 11C, pWBR model, thick trace]. In addition, each type II PVP neuron stopped firing or "paused" during contralaterally directed saccades and vestibular nystagmus quick phases (note vertical arrows in Fig. 11, A and C). The head-velocity sensitivity of type II PVP neurons was obtained during pWBRc using Estimate 1 (mean sample VAF = 0.50 ± 0.1). The example neuron was representative in that its estimated head velocity was reduced by ~21% as compared with pWBR [normalized g_est = 0.79 (sp/s)/(°/s); mean normalized g_est = 0.71 ± 0.06 (sp/s)/(°/s); Fig. 11D, pWBRc estimate, thin trace]. This reduction was comparable to what we observed for type I PVP neurons. However, while type II PVP neurons paused during vestibular nystagmus quick phases and saccades in the "off direction" for eye motion (see contralateral saccade in Fig. 12A, left), they most often (n = 9/14) burst during saccades in the "on direction" (ipsilateral saccade shown in Fig. 12A, right). This latter behavior differed from that of type I PVP neurons. During ipsilaterally directed saccades, the eye-velocity sensitivity (r_sac) of type II PVP neurons was estimated [fr = bias_x + (k_x * eye position) + (r_sac * eye velocity)] and found to be 0.27 ± 0.05 (sp/s)/(°/s).

View larger version (21K): [in this window] [in a new window]   Fig. 11. Activity of an example type II PVP neuron (unit c66_1) during the head-restrained condition. A: the neuron increased its discharge for ipsilaterally directed eye movements and paused for contralaterally directed saccades (vertical arrows). Inset: mean neuronal firing rate was well correlated with horizontal eye position during periods of steady fixation. B: the neuron also increased its discharge during ipsilaterally directed smooth pursuit. A model based on the bias discharge, the eye-position sensitivity, and the eye-velocity sensitivity of the neuron provided a good fit to the neural activity (smooth pursuit model, thick trace). C: passive whole-body rotation was used to characterize the neuron's response to head movements during VOR in the dark (pWBR). A model based on the bias discharge, the eye-position sensitivity, and the head-velocity sensitivity during the compensatory eye movements made during pWBR is superimposed on the firing rate trace. The vertical arrows indicate pauses in activity for contralaterally directed vestibular quick phases. D: unit b39_1 was typical in that its modulation was less during passive whole-body rotation while the monkey cancelled its VOR (pWBRc) by fixating a target that moved with the table (pWBRc estimate, thin trace) than during pWBR (pWBR model, thick trace).

View larger version (34K): [in this window] [in a new window]   Fig. 12. Activity of an example type II PVP neuron (unit cr93_1) during saccades and eye-head gaze shifts. A: this unit was typical in that it paused for contralaterally directed saccades (left) and burst for ipsilaterally directed saccades (right). B: during contralaterally directed gaze shifts of all amplitudes, the neuron paused in activity. C: the neuron burst in activity during both large (left) and small (right) ipsilaterally directed gaze shifts. D: as with type I PVP neurons, the head-velocity sensitivities of type II PVP neurons decreased as a function of amplitude for ipsilaterally directed gaze shifts (compare ) and were always significantly reduced as compared with pWBR ().

Similar to their responses during head-restrained saccades, most (10/13) type II PVP neurons paused for the duration of eye-head gaze shifts in the contralateral direction (Fig. 12B). In addition, most (8/13) neurons burst for the duration of gaze shifts in the ipsilateral direction (Fig. 12C). Recall that the activity of type I PVP neurons was attenuated, in an amplitude-dependent manner, during ipsilaterally directed gaze shifts. We estimated the eye- and head-velocity sensitivity of each type II PVP neuron during ipsilaterally directed gaze shifts using Estimate 2. Even though they burst, the head-velocity sensitivities of type II PVP neurons were significantly reduced for gaze shifts of all amplitudes (Fig. 12D; compare  and ). While the attenuation was not as great as that observed for type I PVP neurons (Fig. 2B), the level of attenuation did significantly increase with the increasing amplitude (R² = 0.83). In addition, the estimated eye-velocity sensitivities were comparable to those estimated during saccades on a neuron-by-neuron basis [sample mean = 0.27 ± 0.05 (sp/s)/(°/s); slope = 0.90, R² = 0.96]. Type II PVP neurons showed similar responses when the monkey rapidly redirected its gaze during all of the behavioral tasks employed. Thus in light of the results that we obtained from type I PVP neurons, we elected to limit our analysis of rapid gaze redirections to saccades and gaze shifts (as described in the preceding text).

Type II PVP neurons behaved similarly to type I PVP neurons whenever the monkey slowly redirected its gaze. The head-velocity sensitivity of the type II PVP neurons was reduced as compared with pWBR (Fig. 13; solid column) when the monkeys cancelled their VOR (pWBRc; Fig. 13; diagonally striped column) and during eye-head gaze pursuit (Fig. 13; vertically striped column). The attenuated response to head velocity was not dependent on whether the head motion was passively or actively generated but rather on the gaze goal of the monkey. This is illustrated by the response of the neurons during simultaneous passive whole-body rotation and voluntary head motion. When gaze was stable, the pWBR model provided a good prediction of the neural firing rate (sample mean VAF = 0.52 ± 0.07) and the head-velocity sensitivity was not attenuated as compared with pWBR (Fig. 13; gray shaded columns). When gaze was redirected the head-velocity sensitivity of the neurons was significantly attenuated (Fig. 13; horizontally striped columns). In addition, during the driving paradigm, where the monkey moved its head and body together in space, the neurons were less responsive when gaze was redirected (Fig. 13; open column) than when gaze was stabilized (Fig. 13; gray shaded column with horizontal stripes). The pWBR model provided a good prediction of the neural activity when gaze was stable during the driving paradigm (sample mean VAF = 0.64 ± 0.14). During ipsilaterally directed rapid gaze redirections (Fig. 13; gray shaded column with diagonal stripes), type II PVP neurons responses were more attenuated than during slow gaze redirection, similar to what was observed for type I PVP neurons (Fig. 10). The influence of afferent inputs from neck muscle proprioceptors on the activity of type II PVP neurons was tested using the same paradigms as described for type I PVP neurons [i.e., passive rotation of the head on the body (Fig. 13; gray shaded column with vertical stripes) and passive rotation of the body under an earth stationary head (data not shown)]. As with type I PVP neurons, the type II PVP neurons tested were not influenced by passive activation (dynamic or static) of the neck proprioceptors.

View larger version (45K): [in this window] [in a new window]   Fig. 13. Summary of type II PVP neuron discharge activity during passive and voluntary head motion. Neuronal responses to head velocity were significantly attenuated when gaze was redirected during both passive and voluntary head motion as compared with pWBR (solid column). Such attenuation occurred during pWBRc (diagonally striped column), eye-head gaze pursuit (vertically striped column; n = 12), and the driving task (open column; n = 4). Similarly, when monkeys made voluntary head-on-body movements during simultaneous passive whole-body rotation, neuronal responses to both the voluntary and passive components of head-in-space motion (compare horizontally striped columns; n = 5) were attenuated. Neurons were attenuated to a greater extent during ipsilaterally directed gaze shifts than during slow gaze redirection (gray shaded column with diagonal stripes). Neurons were not attenuated when the monkey stabilized its gaze during simultaneous passive and voluntary head motion (gray shaded columns), the driving task (gray shaded column with horizontal stripes), and passive head-on-body rotations (gray shaded column with vertical stripes; n = 10). The horizontal dashed lines represent the average normalized head-velocity sensitivities across each condition with the same gaze goal.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gaze stabilization vs. gaze redirection

The main finding of the present study is that the head-velocity-related response of the direct VOR pathways is modulated in a manner that is consistent with the behavioral goal of the animal. Neuronal activity of VOR interneurons (type I PVP neurons) was recorded during a diverse range of vestibular stimuli protocols including: passive whole-body rotation, passive head-on-body rotation, active eye-head gaze shifts, active eye-head gaze pursuit, self-generated whole-body motion (i.e., driving), and active head-on-body movements made while the monkey was passively rotated. Regardless of the stimulation condition, head-velocity-related modulation of type I PVP neurons was comparable whenever monkeys stabilized their gaze relative to space. In contrast, whenever the monkeys' behavioral goal was to redirect their gaze relative to space, type I PVP neuron responses to head motion were significantly reduced. We also found that type II PVP neurons, which are likely to contribute indirectly to the VOR, generally behaved in a similar manner. However, there were some important differences which we consider later in this discussion.

Absence of influence from neck proprioceptive inputs

In the present study, we found that neither static nor dynamic activation of neck propioceptors alone influenced the activity of PVP neurons (Fig. 14). This result is very different from that of a recent study (Gdowski and McCrea 2000) in which it was reported that the majority of type I PVP neurons are sensitive to static head position relative to the trunk (61%) and dynamic neck motion (76%). These investigators noted that their findings might have important implications regarding the COR, which functions to generate compensatory eye movements in response to neck motion. They proposed that because PVP neurons carry neck-related information, these same premotor VOR interneurons might also mediate the COR. On the one hand, the difference between our results and those of McCrea and colleagues is surprising given that neurons were recorded during the same paradigms: passive rotation of the monkey's body under its earth-fixed head and/or passive rotation of its head on its body. On the other hand, an important difference between these two studies is that neurons were characterized in old world (rhesus) monkeys in the present study and in new world (squirrel) monkeys in that of Gdowski and McCrea (2000).

View larger version (27K): [in this window] [in a new window]   Fig. 14. Proposed mechanism mediating the gaze goal dependence of type I PVP neurons. Our findings suggest that neither neck proprioceptive information nor an efference copy of neck motor commands influenced neuron activity. In addition, the monkey's knowledge of self-generated head-in-space motion did not contribute to the attenuation. (Note: ×, hypothetical pathways that have been eliminated.) We propose that an inhibitory signal representing an efference copy of the oculomotor/gaze command mediates the attenuation either at the level of the PVP neurons (A) and/or presynaptically to the PVP neurons (B).

Prior studies have determined that COR gains are small to nonexistent in most species including: rhesus monkeys (Bohmer and Henn 1983; Dichgans et al. 1973), humans (Barlow and Freedman 1980; Bronstein 1992; Bronstein and Hood 1986; Huygen et al. 1991; Jürgens and Mergner 1989), rabbits (Barmack et al. 1981, 1989, 1992; Fuller 1980; Gresty 1976), and cats (Fuller 1980). The results of the present study are consistent with these prior reports; none of our rhesus monkeys had COR gains that differed significantly from zero (Fig. 7, B and C). However, it is interesting to note that squirrel monkeys may be an exception to this general rule. In this species, COR gains in the range of 0.4 have been recently reported (Godwski and McCrea 2000). This marked difference between the COR gain of rhesus and squirrel monkeys is consistent with the apparent difference that neck proprioceptive inputs have on premotor vestibular nuclei neurons for these two species. In addition, prior reports by our laboratory and McCrea and colleagues have revealed an analogous difference regarding the influence of neck proprioceptors on another class of vestibular nuclei neurons, vestibular-only neurons. These neurons, which are thought to contribute to the VCR, appear to carry neck afferent signals in the squirrel monkey (McCrea et al. 1999) but not in the rhesus monkey (Roy and Cullen 2001). In summary, the results of this and previous studies suggest that use of neck proprioceptive inputs by vestibular reflex pathways in rhesus and squirrel monkeys differs greatly.

Absence of influence from neck efference copy

An unresolved question in the vestibular literature is whether or not the gain of the VOR differs for active versus passive head motion. Two previous studies have reported higher VOR gains during active head-on-body motion as compared with pWBR when subjects fixated an earth-fixed target (Jell et al. 1988) and in the dark (Demer et al. 1993). The gain enhancement observed by Demer and colleagues (1993) was attributed to an efference copy of the motor command to the neck. However, there is much accumulated evidence that suggests the VOR gains are comparable during active and passive head motion. First, the gains of the VOR during sinusoidal (0.25-1.0 Hz) passive whole-body rotation and active head-on-body motion were similar (Hanson and Goebel 1998). Second, numerous studies have reported that the gain of the VOR was comparable during passive and active head-on-body motion (Foster et al. 1997; Hanson and Goebel 1998; Pulaski et al. 1981; Santina et al. 1999, 2000; Thurtell et al. 1999). Third, in the present report, we have recorded from the VOR interneurons and show that the head-velocity-related modulation of type I PVP neurons was comparable during pWBR and active head movements made when gaze was stable (Figs. 2C, 5B, and 6B). Finally, our neurophysiological results agree with our observation in the present study that there was no significant difference in the gain of the behavioral VOR during pWBR (Fig. 1C; VOR gain  0.94) and immediately following gaze shifts during the active head motion that occurred once gaze was stable (Fig. 2C; VOR gain  0.98). Thus taken together, the results of behavioral and single-unit recording experiments strongly suggest that an efference copy of the neck motor command does not influence to status of the VOR during active head-on-body motion (Fig. 14).

Our findings appear to be consistent with those of a recent report by Gdowski and McCrea (1999). These investigators recorded the responses of PVP neurons in squirrel monkeys when slow head-on-body movements were made during passive whole-body rotation. They attributed these head-on-body movements to the vestibulocollic reflex and found that neuronal modulation was better related to head-in-space motion than to passive turntable motion. It is probable (although it is not explicitly stated) that the analysis was limited to intervals in which the axis of gaze was stable. Accordingly, these results could be interpreted as further evidence that PVP neurons similarly encode active and passive head motion when gaze is stable.

Influence of knowledge of self-generated motion

Traditionally the vestibular system is associated with generating the reflexes that are crucial for our daily activities, such as stabilizing gaze (Grossman et al. 1988, 1989) and posture (for review, see Peterson and Richmond 1988). However, the role of the vestibular system is not limited to these functions. The development of an accurate spatial representation, proper implementation of navigation, and gaze control involves the interaction between many brain structures that receive vestibular information and in turn project back to the vestibular nuclei. It is possible that these cortical (reviewed in Fukushima 1997) and cerebellar (Voogd et al. 1996) projections could impinge on type I PVP neurons and result in the differential encoding of head velocity. Here, we have examined whether a monkey's knowledge of its self-generated motion modified the head-velocity signals carried by type I PVP neurons. We found that for active head movements made while gaze was stable, the discharge of type I PVP neurons could be accurately predicted by their head-velocity-related response during passive whole-body rotation. This finding was consistent for active head-on-body movements, active movements of the head and body together in space (i.e., the driving paradigm), and combined active and passive head rotations. In all cases, the attenuation of vestibular responses was limited to the specific intervals in which monkeys actively redirected their gaze. We therefore conclude that knowledge of self-motion, per se, does not directly influence the vestibular sensitivity of type I PVP neurons (Fig. 14). Note, we have previously shown that another class of neurons in the vestibular nucleivestibular-only neurons, which are thought to mediate the vestibulocollic reflexare also not directly influenced by the monkey's knowledge of self-generated motion (Roy and Cullen 2001). Thus our results support the hypothesis that a monkey's knowledge of its self-generated head motion relative to space does not alter the processing of vestibular information at the level of the vestibular nuclei.

Convergence of signals on type I PVP neurons

Although the head-velocity-related modulation of type I PVP neurons was significantly attenuated whenever the monkey wanted to redirect its gaze in space, the amount of suppression differed depending on the type of gaze movement. Type I PVP neurons were more attenuated during rapid gaze redirections (Fig. 10, ), which included vestibular quick phases, saccades, eye-head gaze shifts, and head/body-eye gaze shifts (i.e., driving paradigm step target), than during slow gaze redirection (Fig. 10, ), which included VOR cancellation, gaze pursuit, and head/body-eye pursuit (i.e., driving paradigm pursuit target). We propose efference copies of oculomotor/gaze motor commands are responsible for the behaviorally dependent modulation of type I PVP neuronsand as a result, for the status of the VORduring gaze redirection (Fig. 14). Accordingly, the differing levels of suppression result from the different gaze premotor circuitries that generate rapid versus slow gaze redirection.

MECHANISMS FOR VOR SUPPRESSION DURING RAPID GAZE REDIRECTION. During the rapid redirection of gaze (i.e., saccades, vestibular quick phases, and gaze shifts), the brain stem burst generator is active. We have previously proposed that this premotor brain stem circuitry mediates the attenuation of type I PVP neuron responses, which can be observed during each of these behaviors (Fig. 15A) (Roy and Cullen 1998). Burst neurons in the paramedian pontine reticular formation (PPRF) generate a burst in activity to drive the eye during saccades and gaze shifts (Cullen and Guitton 1997). Burst neurons project to type II neurons in the vestibular nucleus (Sasaki and Shimazu 1981), and in turn, type II neurons send an inhibitory projection to type I PVP neurons (Nakao et al. 1982). Because the type II-type I vestibular projection is inhibitory, this pathway would effectively invert the "burst" behavior of burst neurons to create the "pause" in the type I PVP response observed during rapid redirection of gaze.

View larger version (37K): [in this window] [in a new window]   Fig. 15. Possible brain stem premotor circuitries involved in the attenuation of type I PVP neurons during gaze redirection. A: mechanisms for VOR suppression during rapid gaze redirection. Type I PVP neurons receive a strong monosynaptic connection from the ipsilateral vestibular afferents and in turn project directly to extraocular motoneurons. During saccades, vestibular quick phases, and gaze shifts, brain stem burst neurons in the PPRF are active. These neurons are known to project to neurons with type II vestibular responses, which in turn would inhibit the type I PVP neurons. Type II PVP neurons and burst-tonic (BT) neurons are possible candidates for this interneuron. B: mechanisms for VOR suppression during slow gaze redirection. Eye/head (E/H) neurons receive information from the cerebellum during slow gaze redirection and are likely to project to BT neurons. An inhibitory projection from BT neurons could attenuate vestibular responses of type I PVP neurons during gaze pursuit and pWBRc. E/H neurons are modulated during pWBRc, such that an additional head-velocity-related input is required to offset this input on BT neurons, which are not modulated. The role of type II PVP neurons is not clear (?) because their activity during slow gaze redirection is not appropriate to suppress type I PVP neurons.

MECHANISMS FOR VOR SUPPRESSION DURING CANCELLATION AND SLOW GAZE REDIRECTION. Because less is known about the connectivity of the brain stem premotor circuits that mediate smooth pursuit and VOR cancellation than those that mediate saccades, elucidating the mechanism responsible for the attenuation in type I PVP neuron responses during slow gaze redirection is less straightforward. During smooth pursuit eye movements in head-restrained animals, the cerebellar flocculus and paraflocculus send pursuit command signals to neurons in the vestibular nuclei (Lisberger et al. 1994a). Based on their responses during eye movements and head-rotation paradigms, it is generally agreed that these neurons are part of the same population of neurons that have been termed smooth-pursuit and eye-head neurons (Cullen et al. 1993 and Scudder and Fuchs 1992, respectively). Thus for the sake of simplicity, we will refer to them as eye-head (E/H) neurons here. There is no evidence that E/H neurons project to PVP neurons. It is possible that pursuit information reaches the vestibular nuclei indirectly via the BT neurons, which show robust modulation during smooth pursuit (Cullen et al. 1993; McConville et al. 1996; McFarland and Fuchs 1992). A schema in which E/H neurons project to BT neurons and BT neurons, in turn, send inhibitory pursuit signals to type I PVP neurons is shown in Fig. 15B.

Type II PVP neurons

The responses of type II PVP neurons during slow gaze-redirection tasks were not consistent with their playing a primary role in mediating the attenuation of type I PVP neurons if one assumes an inhibitory ipsilateral projection from type II PVP neurons to the type I PVP neurons (as we did in Fig. 15A). For instance during pWBRc, type II PVP neuron responses are attenuated as compared with during pWBR, which means that their modulation is opposite to what would be required to mediate the suppression of type I PVP neurons. This is not only a problem for pWBRc but also for all of the other slow gaze-redirection tasks because type II PVP neuron responses were attenuated (see Fig. 13). Thus while it is conceivable that type II PVP neurons project to type I PVP neurons, their activity does not appear to be integral for the short-term modulation of type I PVP neuron responses across different behavioral conditions. Given the complex interconnections between the vestibular nuclei, the prepositus hypoglossi, and the cerebellum, it is more likely that over the long term they play a role in balancing vestibular function across these structures.

Conclusion

In conclusion, we have shown that the activity of type I and II PVP neurons is modulated in a manner that depends strictly on the current gaze strategy of the monkey. Neuronal discharges were comparable during active and passive head motion whenever the monkey's gaze was stable. Similarly, discharges were attenuated during active and passive head motion when a monkey redirected its gaze. Thus the neuronal responses to head motion are altered in a manner that is consistent with maximizing the VOR gain when the goal is to stabilize gaze and reducing the VOR gain when the behavioral goal is to redirect gaze. We propose that the attenuation of the direct VOR pathways is mediated via inputs from the premotor circuitries that are known to generate saccades and smooth pursuit eye movements in head-restrained animals.