Baclofen, motion sickness susceptibility and the neural basis for velocity storage

Abstract

Reduction of the dominant time constant (T_VOR) of the angular vestibulo-ocular reflex (aVOR) by habituation is associated with a decrease in motion sickness susceptibility. Baclofen, a GABA_b agonist, reduces the time constant of the velocity storage integrator in the aVOR in a dose-dependent manner. The high frequency aVOR gain is unaltered by baclofen. Here we demonstrate that the reduction in T_VOR produced by oral administration of 20 mg of baclofen causes a significant reduction in motion sickness susceptibility, tested with roll while rotating (RWR). These data show that motion sickness susceptibility can be pharmacologically manipulated with a GABA_b agonist and support our conclusion that motion sickness is generated through velocity storage. We also show how baclofen acts on velocity storage at the neural level. A vestibular-plus-saccade (VPS) neuron was recorded in the rostral medial vestibular nucleus (rMVN) of a cynomolgus monkey, an area where we postulate that velocity storage is generated. The cell had a time constant during steps of velocity that was close to that of the T_VOR. After parenteral administration of baclofen, there was a similar decrease in the time constants of the VPS neuron and the T_VOR. This is the first demonstration of the concurrence of unit and aVOR time constants before and after baclofen. The data support the hypothesis that the velocity storage integrator is generated through activity of vestibular-only (VO) and VPS neurons in rMVN and suggest that GABA_b synapses on VO and VPS neurons are likely to be involved in the baclofen-induced reduction in motion sickness susceptibility.

Introduction

Baclofen, 4-amino-3,4-chlorophenyl butanoic acid (Fig. 1A), is a GABA_b agonist that was initially introduced to reduce muscular spasm. However, baclofen also actively modifies the behaviour of the oculomotor and vestibular systems; it abolishes Periodic Alternating Nystagmus (Halmagyi et al., 1980) by reducing the time constant of the velocity storage integrator in the vestibular nuclei (Leigh et al., 1981; Cohen et al., 1987). In particular, there is a dose-dependent reduction in the time constant of optokinetic after nystagmus (T_OKAN) and the angular vestibulo-ocular reflex (T_VOR) in monkeys after intramuscular injection of baclofen (Cohen et al., 1987). The initial jump in eye velocity at the onset and end of rotation was unaffected (Fig. 1C-E). Baclofen similarly affects the human T_VOR (Dai et al., 2006).

Fig. 1.

Baclofen and its effects on the aVOR. (A) Abbreviated chemical structure of baclofen. (B) Reciprocally related plasma levels of baclofen (solid circles) with reduction in human T_VOR (open circles) after single doses of baclofen (Adapted with permission from Faigle et al., 1980 and Dai et al., 2006). (C–E) Per- and post-rotatory nystagmus (top traces) and slow phase velocity of a rhesus monkey (bottom trace) induced by a step of velocity of 60°/s in darkness before (C), 2 h after, (D), and 24 h after (E) an injection of 0.6 mg/kg of baclofen (Adapted with permission from Cohen et al., 1987). (F) Model of the aVOR and of control of motion sickness susceptibility. See text for details. (G, H) Model-based per- and post-rotatory responses of human aVOR before (G) and after (H) ingestion of 20 mg of baclofen. The cupula/endolymph and direct pathway response (cupula) were unaffected, but the velocity storage component in G (arrow), was reduced to produce a shortened T_VOR (Adapted with permission from Dai et al., 2006).

We have proposed that motion sickness induced by rolling the head while rotating, is produced by a disparity between the yaw axis orientation vector of velocity storage and the vector of the slow phase eye velocity generated through velocity storage (Dai et al., 2003). The orientation vector of velocity storage lies close to the axis of gravity (Dai et al., 1991; Raphan and Sturm, 1991; Raphan and Cohen, 2002), while the slow phase velocity vector can deviate substantially from gravity depending on the stimulus (Dai et al., 2003). If the T_VOR is long, then any induced disparity between the eye velocity vector generated through velocity storage and the spatial vertical would persist for a longer time, producing motion sickness more easily. If our formulation is correct, then a reduction in T_VOR should produce a reduction in motion sickness susceptibility. This study presents preliminary results that indicate that this hypothesis is correct. We also present findings that support our previous postulate that velocity storage is produced by the action of vestibular-only (VO) and vestibular-plus-saccade (VPS) cells in the rostral medial and superior vestibular nuclei (Reisine and Raphan, 1992; Yokota et al., 1992).

Methods

Limitations of space preclude a full description of the techniques and methods. For a recent review of models of the aVOR, see Dai et al., 1999, Raphan and Cohen, 2002; for eye movement recording in animals and humans, see Yakushin et al., 1995 and Dai et al., 1999, 2003. For effects of baclofen on the time constant of the aVOR in monkeys and humans, see Cohen et al., 1987 and Dai et al., 2006 and for studies of motion sickness, including techniques for producing motion sickness, see Dai et al., 2003. Informed consent was obtained in the human studies, which were approved by the Institutional Review Board of the Mount Sinai School of Medicine. Adequate protection of animals was maintained under the aegis of the Institutional Animal Care and Use Committee. We first present a simplified model of the aVOR to aid in understanding the effect of baclofen ingestion (Fig. 1F). This is followed by a description of the technique of roll while rotating (RWR), which was used to produce motion sickness in this study (Fig. 2). A critical aspect of the RWR paradigm, initially introduced by Graybiel and associates (Miller and Graybiel, 1973), was that the nystagmus induced by head movement was allowed to complete its course before the next head movement began (Dai et al., 2003). This allowed the full participation of velocity storage in producing motion sickness.

Fig. 2.

Roll while rotating (RWR) paradigm used to elicit motion sickness. (A-1) Pitch (top trace) and yaw (second trace) eye velocities during nystagmus induced by rotation at 138°/s (third trace) about a spatial vertical axis with the head statically tilted 45° throughout the rotation (insert below). Velocities down and to the right are positive according to right hand rule. (A-2) Perception during such rotation (circular arrow) was solely that of rotation about a vertical axis; motion sickness was not induced by this stimulus. (A-3) Reconstructed eye velocity vector of the slow phases of pitch and yaw nystagmus. The vector declined toward zero along an axis parallel to gravity (arrow). (B) Scheme as in (A). When the subject tilted his head to the right while rotating to the left, there were upward pitch and left yaw slow phase velocities. This produced a strong sensation of pitch (circular arrow), and an eye velocity vector that was tilted 59° from the spatial horizontal. The eye velocity vector lay close to the stimulus velocity vector (arrow, B-2). This produced a strong sense of nausea, dizziness, and motion sickness. (C) Model-based simulation of the nystagmus induced by pure rotation to the left about a vertical axis (left, insert above), by tilting the head 45° to the right while rotating (middle, insert above), and by bringing the head back to the upright during rotation (right, insert above). Note that the simulated eye velocities in the middle portion of (C) are close to the actual eye velocities, derived experimentally in (B) (Adapted with permission from Dai et al., 2003).

Results

Model of velocity storage and determination of its parameters

The aVOR comprises “direct” and “indirect” pathways that process the head velocity signal transduced by the semicircular canals (SCC) (Fig. 1F; Raphan et al., 1979; Raphan and Cohen, 2002). The direct pathway, which is controlled by a gain g₁, has a fast response time that stabilizes the retina in space against rapid angular head perturbations. The dynamics of the direct pathway are dependent on a three-neuron arc (Lorente de Nó, 1932), which has a minimum latency of transmission between the SCC and the eye muscles of about 2.5 ms (Cohen and Suzuki, 1963). Consequently, the eyes can respond to impulses of angular acceleration in 12 ms (Lisberger and Pavelko, 1988).

The indirect pathway is controlled by an integrator whose coupling from the SCC is represented by g₀, and whose dynamics are represented by the feedback, h. The output is realized in velocity storage neurons Vn, which project both to the oculomotor system and the flocculus (Zhang et al., 1993). The feedback loop is controlled by the nodulus and uvula of the vestibulo-cerebellum, which determines the time constant (1/h) and the orientation of velocity storage that tends to align the velocity vector of nystagmus to the spatial vertical (Dai et al., 1991; Raphan and Sturm, 1991; Wearne et al., 1998). The nodulus also controls habituation, which produces a reduction in the time constant of the velocity storage integrator in response to repetitive exposure to rotation (Cohen et al., 1992). A more complete description is presented elsewhere (Raphan et al., 1979; Raphan and Cohen, 2002). We have proposed that the orientation vector generated by the nodulus/uvula and the eye velocity vector produced by velocity storage is processed by the Orientation Comparator (Fig. 1F). If there is a discrepancy between the two vectors, the autonomic system is activated to produce a buildup of motion sickness (Dai et al., 2003).

For constant velocity rotation about a yaw axis, per- and post-rotatory nystagmus is induced at the onset and end of rotation, shown here for the monkey (Fig. 1C-E; Cohen et al., 1987). The direct pathway response, which is dependent on the dynamics of the cupula/endolymph system, is shown by the solid falling lines labelled “Cupula” in Fig. 1G, H (arrows). The velocity storage or indirect pathway response is shown by the solid lines and arrows labeled “Velocity Storage” in Fig. 1G, H. It has two modes: a rising component due to the input from the SCC and a slower response due to activation of the velocity storage integrator. The induced activity falls with the time constant of the integrator, which is governed by the parameter h, whose value is controlled by the nodulus (Fig. 1F). The signals from the direct and indirect pathways sum in the model at Vn (Fig. 1F) to generate the eye velocity command that produces the slow phase eye velocity and triggers the quick phases of nystagmus. As shown by Fig. 1G, after the first 10 s, there is essentially no input from the labyrinth and the magnitude and decline in the slow phase eye velocity is entirely dependent on the activation of velocity storage. Thus, the time constant of per- and post-rotatory nystagmus in subjects with time constants longer than 8–10 s primarily represents activity in velocity storage (Dai et al., 1999).

Effects of baclofen on the aVOR

Following the injection of baclofen in monkeys, T_VOR was reduced but recovered within 24 h (Fig. 1C-E). A similar decline in T_VOR occurs in humans after ingestion of 20 mg of baclofen (Figs. 1B and 3A, C). In addition to the changes in velocity storage time constant, there were also changes in the coupling of vestibular input to the velocity storage mechanism in the vestibular nuclei (g₀; Dai et al., 2006).The reductions in T_VOR in both humans and monkeys were closely correlated with the rise in the levels of plasma baclofen in humans, determined separately (Fig. 1B; Faigle et al., 1980). Changes in T_VOR over 4 h in six subjects after administration of 20 mg and 30 mg of baclofen followed approximately the same time course (Fig. 3C). This indicates that 20 mg is above threshold for inducing the reduction in T_VOR.

Fig. 3.

Changes in motion sickness susceptibility tested with RWR before and after baclofen. (A) Changes in T_VOR during testing with 138°/s steps of velocity in dark before (top graph) and after ingestion of 20mg of baclofen (bottom graph, solid circles). There was little habituation over the 6 h period of testing (top graph), but the subject’s T_VOR fell from 14 s to 9 s after baclofen over an equivalent period of testing a month later (bottom graph). (B) Number of head movements during RWR before baclofen (left bar), 2 h after baclofen (centre bar), and 5 months after baclofen (right bar). The subject could only initially make 18 head movements without reaching a motion sickness score of 20, but was able to make 50 head movements 2 h after receiving baclofen. Five months later, he was only able to make 23 head movements without baclofen before reaching a motion sickness score of 20 (about to vomit). (C) Comparison of T_VOR during rotational testing at 60°/s without medication (open circles), and up to 4 h after receiving 20 mg (filled circles) and 30 mg (filled squares) of baclofen. The grey areas show the 95% confidence intervals. After baclofen, T_VOR fell and remained depressed over the 4 h of testing. There was no difference in the reduction in T_VOR produced by doses of 20mg and 30mg of baclofen. (Adapted with permission from Dai et al., 2006). (D) Increase in motion sickness score, based on steady-state levels of nausea, before (filled circles) and after baclofen (open circles). The ordinate is Head Movement Number, and the abscissa is the Motion Sickness Score. If subjects did not reach a motion sickness score of 20, the test was terminated after 50 head movements. In each subject, the rise in motion sickness was slower after receiving 20 mg of baclofen. The T_VOR (Tc’s) before and after receiving baclofen are shown adjacent to each of the graphs.

A consistent finding has been that there is no change in the gain of the aVOR following the administration of baclofen (Fig. 1C-E, G, H; Cohen et al., 1987; Dai et al., 2006). From this, we conclude that baclofen acts primarily on the velocity storage component of the aVOR and has little effect on the direct pathway (Dai et al., 2006). Side effects were minimal at 0.3 mg/kg, being predominantly drowsiness. At considerably higher doses in monkeys (≈–2 mg/kg), the time constant of the final common “neural integrator,” which is utilized in producing positions of fixation and in linearizing the slow phases of optokinetic and vestibular nystagmus, also becomes shortened (Yakushin, unpublished data). However, at lower dose levels, baclofen is a convenient way to reduce the time constant of the aVOR, and as we will demonstrate, to reduce motion sickness susceptibility.

Production of motion sickness by roll while rotating (RWR)

Rolling the head while rotating about a spatial vertical axis, activates both the vertical and horizontal SCC. This produces pitch and yaw nystagmus, a sense of pitching forward or backward, and nausea, dizziness, and vertigo. It is a potent stimulus for generating motion sickness (see Dai et al., 2003, for review). Model-generated responses show the direction and magnitude of the expected nystagmus (Fig. 2C). Rotation to the left about a vertical axis induces horizontal nystagmus to the right, which decays to zero as the rotation continues (Fig. 2C, left, Yaw Eye Vel). When the head is rolled to the right (insert above), the vertical canals move into and the lateral canals out of the plane of rotation. This produces upward per-rotatory pitch and leftward post-rotatory yaw eye velocity. The simulated velocity vector of the induced nystagmus is shown by the arrow in the middle portion of Fig. 2C (middle, arrow).

The pitch per- and yaw post-rotatory responses in the vertical and lateral canals produced in the model by head roll, were present experimentally (Fig. 2B, 1st and 2nd traces). The velocity vector of the stimulus in space was 67° (Fig. 2B-2). In response, the eye velocity vector (Fig. 2B-3) declined toward zero along a trajectory that was tilted from the vertical by 59°, similar to that predicted by the model (Fig. 2C, arrow). Concurrently, there is a strong sense of pitching forward (Fig. 2B-2, rotatory arrow) and a sudden sense of disorientation and dizziness. The increase of nausea associated with the roll head movement is momentary, lasting for about 30 s and then declines to a steady state level.

When the head is brought back to the upright (Fig. 2C, right), the direction of the nystagmus reverses, producing a downward pitch post-rotatory and a leftward yaw per-rotatory response. Again there is a transient increase in motion sickness, followed by a decline to a steady state level of nausea that is generally more intense than after the preceding head movement. As the sequence of transient and steady state increments of nausea continue, the level of steady state nausea increases after each head movement until a state of imminent emesis is reached, i.e., a motion sickness score of 20, at which point, the test is terminated. The test is also terminated if a score of 20 is not reached after 50 head movements (A more complete description of the paradigm is given in Dai et al., 2003).

The importance of the disparity between the eye velocity vector and the orientation vector of velocity storage for production of motion sickness is shown in Fig. 2A. If the head is simply held in a tilted position before and during rotation (Fig. 2A-1), horizontal and vertical slow phase velocities are generated. However, in contrast to the situation in Fig. 2B, C, eye velocity decays along a straight line close to the direction of the acceleration of gravity and the orientation vector of velocity storage (Fig. 2A-2; Dai et al., 1991; Raphan et al., 1992). Perceptually, subjects feel as if they are simply rotating around a vertical axis (rotatory arrow, Fig. 2A-2), and such stimulation did not cause motion sickness in any of the six subjects that were repeatedly tested with this paradigm.

Effects of baclofen on motion sickness susceptibility

An important finding from the study of Dai et al. (2003), was that the T_VOR declined concurrently with the ability of subjects to make more head movements during repeated testing. That is, the two were reciprocally related. This led to the current hypothesis that motion sickness susceptibility could be reduced if there was a pharmacological reduction in T_VOR. Subjects were first tested with RWR to determine their baseline motion sickness susceptibility. A week later, subjects received a single dose of 20 mg of baclofen and were retested with RWR an hour later. Results from one subject are shown in Fig. 3A, B. He made 18 head movements during the pretest before developing severe motion sickness (Fig. 3B, left). Following administration of 20mg of baclofen, his T_VOR dropped from 14 s to 9 s (Fig. 3A), and he was able to make 50 head movements without reaching a motion sickness score of 20 (Fig. 3B, centre). Five months later when effects of baclofen and habituation had disappeared, he could make only 25 head movements during RWR before reaching emesis.

The buildup of the steady state level of motion sickness in four other subjects is shown in Fig. 3D. The T_VOR’s of these subjects before and after baclofen are shown by the number adjacent to the individual curves. Two subjects (D-1 and D-4) reached a motion sickness score of 20 (imminent emesis) both with and without baclofen, but the buildup was slower after receiving the drug. One subject (D-3) never reached a score of 20 either with or without baclofen, but the buildup of symptoms was slower after baclofen. A fourth subject (D-2) had a motion sickness score of 20 after 26 head movements before receiving the drug, but was able to make 50 head movements after receiving baclofen and only reached a score of motion sickness score of 6. Thus, 20 mg of baclofen, which shortened the T_VOR in all four subjects, was effective in reducing their motion sickness susceptibility.

Neural basis for changes in velocity storage

There is evidence that several of the salient characteristics of velocity storage are embedded in VO and vestibular-plus-saccade (VPS) neurons located in rostral medial vestibular nucleus (rMVN) and SVN (Reisine and Raphan, 1992; Yokota et al., 1992; Holstein et al., 1999). We therefore questioned whether the temporal responses of these types of neurons to optokinetic and vestibular stimuli would be modified by the administration of baclofen. A typical VPS neuron did not respond to optokinetic stimulation at frequencies of 0.1 Hz or higher (Fig. 4A), and had variable responses to a frequency of 0.05 (Fig. 4B). There was a robust response to optokinetic stimulation at 0.02 Hz, however (Fig. 4C). The cell had rapid changes in activity at the onset of constant velocity rotation, followed by slower change in firing rate as the rotation continued (Fig. 4D). This neuron also coded direct pathway activity, and rapid changes in firing rate occurred both at the onset and end of rotation that were approximately linearly related to the rotational velocity (Fig. 4E).

Fig. 4.

Graphs of vestibular-plus-saccade neuron and aVOR T_VOR before (A-D) and after (F) parenteral injection of 0.2 mg/kg baclofen in a cynomolgus monkey. (A-C) Unit activity (top trace), desaccaded yaw eye velocity (middle trace), and OKN velocity (lower trace). The optokinetic stimulus was oscillated at 0.1 Hz (A), 0.05 Hz (B), and 0.02 Hz (C). There was some modulation in unit firing at 0.05 Hz, but it became robust at 0.02 Hz. (D, F) Comparison of unit activity (top trace) and T_VOR (2nd trace) before (D) and after (F) injection of baclofen. The velocity of the rotator is shown in the 3rd trace. The time constants of both the unit activity and the T_VOR, were comparable before (D) and after (F) baclofen. (E) Increase and decrease in firing frequency of the unit in response to rotation at velocities from 60°/s to 180°/s in darkness before and after baclofen. Abscissa, stimulus velocity; ordinate, increases (excitatory) and decreases (inhibitory) in firing frequency. There was no effect of baclofen on the initial response to rotation in either the excitatory or inhibitory directions. (G) Comparison of time constants of the aVOR and of unit activity before and after injection of baclofen. There was an increase in the time constants of both the unit activity and eye velocity before baclofen (open circles) in response to increases in stimlulus velocity (R = 0.780). After baclofen (solid circles), the time constants were clustered at lower values.

There was a close correspondence between the unit time constants and the T_VOR at all tested velocities. The T_VOR of the per- and post-rotatory nystagmus shown in Fig. 4D were 20 s and 27 s, respectively (2nd trace), in response to constant velocity rotation at 180°/s (bottom trace). The corresponding time constants of the unit activity associated with this rotation were 23 s and 21 s, respectively (top trace). The close association of the unit and aVOR time constants over trials from 30°/s–180°/s is demonstrated in Fig. 4G by the open circles. After parenteral (IM) administration of 6mg of baclofen (1.5 mg/kg), the time constants of the per- and post-rotatory responses fell to 9 s and 11 s (Fig. 4F, 2nd trace) for the same 180°/s rotation (bottom trace), while the time constants of the associated unit responses decreased to 8 s and 12 s, respectively (top trace). This parallel reduction in time constants was also present in the repeated trials at the different velocities, as shown by the filled symbols in Fig. 4G.

Thus, there was a close correlation of the unit and the nystagmus time constants both before and after administration of baclofen. The changes in firing rate associated with the onset and end of rotation were unaffected by the administration of baclofen (Fig. 4E). This is strong support for the hypothesis that baclofen does not affect the direct pathway response, only the response generated by the velocity storage integrator.

Discussion

We have shown here that oral administration of baclofen, which shortened the aVOR time constant, reduced motion sickness susceptibility. We have further demonstrated that the reduction in T_VOR produced by baclofen was closely paralleled by a reduction in the time constant of a VPS neuron lying in rMVN. Consistent with the postulate that baclofen only affects the indirect and not the direct pathway, the high frequency responses of both the aVOR and of the VPS neuron were unaltered. This study also has clinical implications. An average dosage of ≈0.3 mg/kg, produced by ingestion of 20 mg of baclofen, which had minimal side effects, was effective in reducing T_VOR by 27% on average, in decreasing the coupling gain of efferent vestibular information to velocity storage (g₀) by 24% (Dai et al., 2006), and in reducing motion sickness susceptibility. From the data of Faigle et al. (1980) in human and of Cohen et al. (1987) in monkeys, the reduction in T_VOR and presumably in motion sickness susceptibility would last for up to 8 h–10 h. Thus, this or similar GABA_b agonists are likely to be effective agents to reduce motion sickness susceptibility.

Implications of neural activity associated with velocity storage

It has been known since the pioneering studies of Waespe and Henn (1977) that the time constants of VO and VPS neurons reflect the time constants of OKAN and of per- and post-rotatory nystagmus, i.e., of velocity storage (Cohen et al., 1987; Raphan et al., 1979). This is the first demonstration, however, that a VPS neuron changed its time constant in parallel with the changes in T_VOR after administration of baclofen. This provides additional evidence that velocity storage is coded in the activity of VPS neurons, and raises the possibility that such neurons could be the neural basis for the effect of baclofen on motion sickness susceptibility. The structural basis for this effect and for the associated changes in g₀ (Dai et al., 2006) may come from the immunocytochemical studies of Holstein et al. (1992a, b) that have demonstrated baclofen-sensitive GABA_b receptors in rMVN. Their ultrastructural studies have further shown that such inhibition is mediated by both axo-dendritic (post-synaptic) and axo-axonic (pre-synaptic) contacts (Holstein et al., 1999). Presumably, activation of axo-dendritic synapses could produce a reduction in membrane resistance that would reduce periods of activation of such neurons, resulting in a reduction of T_VOR.

Activity in the direct aVOR pathway is also coded in the firing rates of VO and VPS neurons, but was not affected by baclofen (Fig. 4E). The resting discharge was also not altered. If the reduction in activity related to rotation in the “off-direction” was produced by inhibition, these neurons must have different synaptic inputs and inhibitory transmitters to modulate activity in the direct and indirect pathways. From our results, GABA_b is utilized to inhibit the velocity storage component, but another inhibitory transmitter, possibly GABA_a, may be utilized for inhibition of neurons in the direct pathway.

Implications for aVOR structure

The aVOR model of Fig. 1F shows how baclofen could affect the time constant of the aVOR and motion sickness susceptibility. A key element is the control of spatial orientation and temporal properties of velocity storage by the nodulus and uvula, which set the feedback parameter, h, which controls its time constant (Wearne et al., 1998). Whether baclofen also modifies the yaw axis spatial orientation vector of velocity storage is not clear, but it clearly shortens the velocity storage time constant. Thus, any discrepancy between the orientation vector and the actual eye velocity generated by velocity storage is reduced by baclofen, thereby slowing the motion sickness buildup that accompanies each head tilt (Fig. 3D). Thus, more head movements can be made before a critical level of motion sickness buildup is reached.

In the model of Fig. 1F, only the temporal and orientation properties of velocity storage would impact the generation of motion sickness. Yet, VO and VPS neurons reflect both the activity of velocity storage and the direct pathway (Waespe and Henn, 1977; Reisine and Raphan, 1992). This could possibly indicate that the VO and VPS neurons lie beyond the summing junction of the output of the velocity storage integrator, Vn, where the direct and indirect pathways join to form the Eye Velocity Command. It is possible that there are specific neuron classes in the vestibular nuclei that have not been encountered, which only code activity of the velocity storage integrator. If such neurons do not exist, however, it could imply that there is filtering of the signals coming from the vestibular nuclei so that eye velocity commands are generated that only reflect the low frequency velocity storage component. Alternatively, since baclofen does not affect the direct pathway response of the VPS neurons, velocity storage might be an emergent network property of VO and VPS neurons (Reisine and Raphan, 1992) and not a specific property of individual VO or VPS cells. Regardless, the results support the importance of velocity storage as a major factor in the production of motion sickness.

Abbreviations

aVOR

angular vestibulo-ocular reflex

g ₀

indirect pathway gain

g ₁

direct pathway gain

GABA_a, GABA_b

gamma amino butyric acid a and b

IM

parenteral (intramuscular) injection

OKAN

optokinetic after nystagmus

rMVN

rostral medial vestibular nucleus

RWR

roll while rotating

T _OKAN

time constant of optokinetic after nystagmus

T _VOR

time constant of the angular vestibulo-ocular reflex

VO

vestibular-only neuron

VPS

vestibular-plus-saccade neuron