 |
|||||
|
|||||
|
|||||
|
|||||
|
|||||
Previous Article | Next Article
The Journal of Neuroscience, November 1, 1998, 18(21):9112-9129
Neural Learning Rules for the Vestibulo-Ocular Reflex
Jennifer L. Raymond and Stephen G. Lisberger Howard Hughes Medical Institute, Department of Physiology and W. M. Keck Foundation Center for Integrative Neuroscience, University of California, San Francisco, California 94143
ABSTRACT
Top
Abstract
Introduction
Mechanisms for the induction of motor learning in the vestibulo-ocular reflex (VOR) were evaluated by recording the patterns of neural activity elicited in the cerebellum by a range of stimuli that induce learning. Patterns of climbing-fiber, vestibular, and Purkinje cell simple-spike signals were examined during sinusoidal head movement paired with visual image movement at stimulus frequencies from 0.5 to 10 Hz. A comparison of simple-spike and vestibular signals contained the information required to guide learning only at low stimulus frequencies, and a comparison of climbing-fiber and simple-spike signals contained the information required to guide learning only at high stimulus frequencies. Learning could be guided by comparison of climbing-fiber and vestibular signals at all stimulus frequencies tested, but only if climbing fiber responses were compared with the vestibular signals present 100 msec earlier. Computational analysis demonstrated that this conclusion is valid even if there is a broad range of vestibular signals at the site of plasticity. Simulations also indicated that the comparison of vestibular and climbing-fiber signals across the 100 msec delay must be implemented by a subcellular "eligibility" trace rather than by neural circuits that delay the vestibular inputs to the site of plasticity. The results suggest two alternative accounts of learning in the VOR. Either there are multiple mechanisms of learning that use different combinations of neural signals to drive plasticity, or there is a single mechanism tuned to climbing-fiber activity that follows activity in vestibular pathways by ~100 msec.
Key words: vestibulo-ocular reflex; plasticity; horizontal gaze velocity Purkinje cells; climbing fibers; parallel fibers; cerebellar LTD
INTRODUCTION
A major goal in neuroscience is to link systems-level analyses of learning with cellular analyses of plasticity. At the interface of these two levels of inquiry is the question of what patterns of neural activity are necessary and sufficient to induce synaptic plasticity in the awake behaving animal. Such neural signals must transduce the sensory stimuli that guide learning into the cellular changes that encode memory. In the present study, we ask what patterns of neural activity drive plasticity in vivo by examining the neural signals present during stimuli that induce motor learning in the vestibulo-ocular reflex (VOR).
The VOR stabilizes images on the retina by causing eye rotation in the opposite direction to head turns. Motor learning calibrates the VOR by modifying the amplitude of the reflex whenever retinal image motion is associated persistently with head turns (Gonshor and Melvill Jones, 1973
; Ito et al., 1974
Learning in the VOR is associative: it depends on the pairing of head turns and image motion. Therefore, one would expect the information that guides learning in the VOR to be carried in the correlation between two or more neural signals. Three likely candidates for the neural signals that guide learning in the VOR are the activity in vestibular pathways, activity in climbing fibers carrying visual signals, and the simple-spike activity in Purkinje cells carrying visual, vestibular, and eye movement signals (see Figs. 2, 4, left). All three of these neural signals are present at the putative sites of plasticity in the circuit for the VOR, which are in the floccular complex of the cerebellar cortex and in the vestibular nuclei (for review and citations to the relevant literature, see duLac et al., 1995
In the present study, we constrain hypotheses regarding the neural signals that guide cerebellum-dependent learning by examining the patterns of climbing-fiber, simple-spike, and vestibular signals present during a range of stimuli that induce learned decreases or increases in the gain of the VOR. We focus on a well studied subclass of Purkinje cells called the horizontal gaze velocity Purkinje cells (HGVPs) (Lisberger and Fuchs, 1978a
MATERIALS AND METHODS
Experiments were conducted on two male rhesus monkeys that had been trained to perform a visual fixation task (Wurtz, 1969
During experiments, each monkey sat in a specially designed primate chair, to which his implanted head holder was secured. Platinum-iridium electrodes were used to make recordings from Purkinje cells in the floccular complex of the cerebellum (flocculus and ventral paraflocculus) (Gerrits and Voogd, 1989
After a Purkinje cell was isolated, it was first characterized by its responses (1) during the smooth pursuit eye movements evoked by sinusoidal motion of a small visual target along a horizontal or vertical axis at 0.5 Hz, ±31.4°/sec (position amplitude ±10°) and (2) as the monkey canceled his VOR by tracking a spot that moved exactly with sinusoidal head rotation about a vertical axis at 0.5 Hz, ±31.4°/sec. For these initial behavioral conditions, the visual stimulus was a small spot subtending 0.5° of visual angle. The present work focuses on HGVPs, a well studied class of Purkinje cells in the floccular complex (Lisberger and Fuchs, 1978a
Recordings were made from HGVPs under conditions that had been shown in a previous study to cause learning in the VOR (Raymond and Lisberger, 1996
For each frequency of sinusoidal vestibular stimulation, ×0 and ×2 stimuli were alternated, and each stimulus was presented for 60 to 120 sec. Recordings were made when the gain of the VOR was close to 1.0, and the ×0 and ×2 stimuli were not presented for long enough to cause measurable changes in the gain of the VOR in the dark. Thus, Purkinje cell responses were recorded under conditions that cause learning but were not followed as the gain of the VOR was modified. To maintain a constant level of alertness and to keep the visual stimulus approximately centered in the visual field during vestibular stimulation, monkeys were rewarded at intervals of 1.5-4 sec for keeping their gaze within ±10° of the spot in the center of the visual stimulus. This reward contingency was the same as that used previously to study behavioral changes in the VOR (Raymond and Lisberger, 1996
Electrodes were introduced daily and driven by a hydraulic microdrive through the cerebral cortex toward the cerebellum. Entry into the cerebellum was recognized by a large increase in background activity and the presence of the complex spikes of Purkinje cells. Entry into the floccular complex was signaled by the presence of background activity related to eye movements and confirmed by the location relative to landmarks such as the vestibular nerve and bone. We did not perform histology to verify the location of the electrodes, because HGVPs are known to localize to the floccular complex (Lisberger and Fuchs, 1978a
The responses of individual Purkinje cells were isolated by careful movements of the electrode and then followed for up to 1 hour. Voltages related to eye position, eye velocity, head velocity, and visual stimulus position were recorded during the experiment at 500 Hz/channel. An eye velocity signal was obtained by using an analog circuit to differentiate the eye position output from the eye coil electronics, and the head velocity signal was obtained from a tachometer attached to the shaft of the turntable. The simple-spike activity of Purkinje cells was triggered with a hardware window discriminator, and the times of the resulting pulses were recorded to the nearest 10 µsec. In addition, unit activity was sampled at 50 kHz, and off-line spike sorting with time and amplitude windows was used to discriminate complex spikes and record their time of occurrence to the nearest 1 msec. Of our total sample of 54 HGVPs, complex spikes were analyzed for the 26 in which the complex-spike wave form could be clearly and reliably differentiated from the simple-spike wave form. In 15 HGVPs, the complex spike could be seen or heard, but not discriminated with confidence from the simple spikes in all records. In the remaining 13 Purkinje cells, complex spikes were not observed, but the cells were deemed likely to be Purkinje cells because of their simple-spike wave form, their irregular simple-spike firing rate, the ability to record the cells through several hundred micrometers of cerebellum, and, in some cases, their characteristic injury discharge at the end of the recording.
The data were analyzed after the experiment by aligning the records on the negative-to-positive zero crossings of sinusoidal head velocity or visual stimulus position. For ×0 and ×2 stimuli, tracking was not a criterion, so that all stimulus cycles were included in the analysis. For pursuit and cancellation of the VOR at 0.5 Hz, only cycles with good tracking were included in the analysis. Each cycle was divided into 64 equal-length bins, and head velocity, simple-spike firing rate, and complex-spike firing rate were averaged. The amplitude of modulation and phase of the responses were estimated as the amplitude and phase of the fundamental components provided by Fourier analysis of the averages.
RESULTS
To evaluate which neural signals guide the induction of learning in the VOR, we compared the patterns of neural activity present at the putative sites of plasticity during stimuli that induce learned decreases and increases in the gain of the VOR (Fig. 1). Conditions that induce a learned decrease in gain were created by pairing a vestibular stimulus with a visual stimulus that moved exactly with the head (×0 stimulus). Conditions that induce a learned increase in gain were created by pairing a vestibular stimulus with a visual stimulus that moved exactly opposite to the head (×2 stimulus). We performed a pairwise analysis of the simple-spike, climbing-fiber, and vestibular signals present during ×0 and ×2 stimuli to determine whether each pair of signals contained the information required to guide learning.
View larger version (21K):[in this window]
[in a new window]
Figure 1. Stimuli that induce learned decreases (×0, A) and increases (×2, B) in the gain of the VOR. From top to bottom, the traces are eye velocity with respect to the orbit, angular head velocity in space, visual stimulus velocity in space, and gaze velocity in space. Gaze velocity was computed as the sum of head velocity in space plus eye velocity in the orbit. In all traces, upward deflections represent leftward position or velocity (L); downward deflections represent rightward position or velocity (R). The brief deflections in the eye and gaze velocity traces are caused by saccadic eye movements; their amplitudes have been cropped. The frequency of the stimuli is 0.5 Hz.
We used two criteria for evaluating whether a particular combination of neural signals could provide suitable guidance for the cellular mechanisms of learning in the VOR. First, signals that guide learning must discriminate ×0 stimuli, which decrease the gain of the VOR, from ×2 stimuli, which increase the gain of the VOR. Patterns of neural activity that are the same during both stimulus configurations would contain no information about whether the gain should decrease or increase and therefore could not be responsible for the opposite learned changes in the VOR induced by the ×0 and ×2 stimuli. This criterion is illustrated by previous analyses of the correlation between the vestibular stimulus and simple-spike activity in HGVPs (Lisberger and Fuchs, 1978a
In the present paper, we applied an additional criterion: if there is a single mechanism that mediates learning in the VOR, then a single pair of neural signals should evince similar patterns of activity during all stimuli that induce similar learned changes in the VOR. For example, ×0 stimuli are effective at inducing learned decreases in the gain of the VOR, and ×2 stimuli are effective at inducing learned increases in the gain of the VOR for sinusoidal stimulus frequencies up to at least 5 Hz (Raymond and Lisberger, 1996
We examined three pairs of neural signals: simple-spike activity in Purkinje cells versus vestibular signals; climbing-fiber activity versus vestibular signals; and simple-spike versus climbing-fiber activity. We were able to evaluate all three pairs simultaneously by recording from Purkinje cells under conditions in which we controlled the vestibular inputs. Simple-spike activity was recorded directly from the Purkinje cells. Climbing-fiber activity was recorded by isolating complex spikes from recordings of Purkinje cells, because complex spikes are driven in a one-to-one manner by spikes in the climbing-fiber input to the Purkinje cell. We did not record the vestibular inputs to either the brainstem or cerebellar site of plasticity directly. Because we controlled the vestibular stimulus, however, we could be confident that these inputs were identical for ×0 and ×2 stimuli at a particular frequency. As a result, comparison of Purkinje cell complex-spike or simple-spike discharge with the head velocity stimulus at a particular frequency reveals whether learning could be guided by correlating climbing-fiber responses or simple-spike activity in Purkinje cells with the activity of any set of vestibular neurons. Here, we present the results from the HGVP subclass of Purkinje cells (Lisberger and Fuchs, 1978a
Correlation of Purkinje cell simple-spike activity with the vestibular stimulus during learning
The histograms in Figure 2 show the simple-spike activity in a typical HGVP during stimuli that, if prolonged, would have induced learned changes in the gain of the VOR. All four stimuli caused modulation of the simple-spike activity around a fairly high-average firing rate of ~70-80 spikes/sec. Because of the high spontaneous firing rate of the HGVPs, we analyze the modulation of simple-spike activity rather than absolute levels of activity. At 0.5 Hz, the timing of elevated simple-spike activity relative to the vestibular stimulus discriminated the ×0 stimulus configuration from the ×2 stimulus configuration. When a 0.5 Hz vestibular stimulus was paired with ×0 visual stimulus motion, peak simple-spike activity in the HGVP coincided with ipsiversive head motion (Fig. 2A, vertical dashed line, Ipsi). When the same vestibular stimulus was paired with ×2 visual stimulus motion, peak simple-spike activity coincided with contraversive head motion (Fig 2C, vertical dashed line, Contra). Simple-spike activity also was modulated during the 5 Hz stimuli. However, at this frequency, the relative timing of the vestibular stimulus and simple-spike activity in the HGVP failed to discriminate the ×0 stimulus from the ×2 stimulus. Simple-spike activity peaked during contraversive head velocity, regardless of whether the stimulus induced a decrease (Fig. 2B, ×0) or increase (Fig. 2D, ×2) in the gain of the VOR.
View larger version (27K):[in this window]
[in a new window]
Figure 2. Histograms showing the simple-spike activity recorded in a representative Purkinje cell during stimuli that induce learned decreases (×0, A, B) and increases (×2, C, D) in the gain of the VOR. Head velocity, Angular head velocity in the horizontal plane. Vertical dashed lines mark peak contraversive and ipsiversive head velocity. Note the different time scales in the left (A, C) and right (B, D) panels, which show data for sinusoidal stimuli at 0.5 and 5 Hz, respectively. Twenty to 500 stimulus cycles were averaged to obtain each histogram. The simplified circuit diagram on the left highlights the loci of vestibular and Purkinje cell (PC) simple-spike signals in the circuit for the VOR.
Figure 3 summarizes the correlation of the vestibular stimulus with simple-spike activity recorded in the entire sample of 54 HGVPs. Responses are plotted in polar coordinates, at a distance from the origin corresponding to the amplitude of response modulation and an angle corresponding to the phase of peak simple-spike activity relative to the vestibular stimulus. Simple-spike responses in phase with peak ipsiversive head velocity are plotted in Figure 3 to the right of the origin, responses in phase with peak contraversive head velocity are plotted to the left of the origin, and clockwise rotation around the graph represents increased phase lead. Thus, in these polar plots, the phase reflects the relative timing of the simple-spike and vestibular signals. The key question is whether the responses to ×0 and ×2 stimuli plot in separate parts of each graph or whether the responses during ×0 and ×2 stimuli are similar and plot together.
View larger version (26K):[in this window]
[in a new window]
Figure 3. Summary of Purkinje cell simple-spike responses, plotted relative to the vestibular stimulus. Each plot compares responses during ×0 stimuli and ×2 stimuli at the single frequency indicated in the bottom right quadrant. Each point represents the simple-spike activity recorded in a single Purkinje cell, plotted in polar coordinates with distance from the origin corresponding to the amplitude of response modulation and an angle corresponding to the phase shift between peak simple-spike activity and head velocity. Responses in phase with peak ipsiversive head velocity are plotted to the right of the origin, responses in phase with peak contraversive head velocity are plotted to the left of the origin, and clockwise rotation around the graph represents increased phase lead. An individual Purkinje cell contributed two symbols for each stimulus frequency: a + (monkey D) or X (monkey E) symbol for the ×0 stimulus, and a filled square (monkey D) or filled triangle (monkey E) for the ×2 stimulus. The plots in A-D are on the same scale, with inner and outer circles representing 10 simple spikes/sec (SS/s) and 50 simple spikes/sec, respectively.
At 0.5 Hz (Fig. 3A) and 2 Hz (Fig. 3B), the patterns of simple-spike and vestibular signals present during the ×0 stimuli (Fig. 3, + and X symbols) were clearly different from the patterns of signals present during the ×2 stimuli (Fig. 3, filled symbols). Responses to the ×0 and ×2 stimuli form distinguishable, almost nonoverlapping populations that plot on opposite sides of each graph. Thus, the timing of the simple spikes relative to the vestibular stimulus can discriminate ×0 from ×2 stimuli at these low frequencies. In contrast, at 5 Hz (Fig. 3C) and 10 Hz (Fig. 3D), the patterns of simple-spike and vestibular signals present during the ×0 stimulus configuration were quite similar to those present during the ×2 stimulus configuration. The polar plots in Figure 3 show extensive overlap in the response populations, even for the 5 Hz stimuli, which produced simple-spike responses of amplitudes similar to those produced by lower-stimulus frequencies. This indicates that the relative timing of simple-spike and vestibular signals cannot distinguish the ×0 stimuli from the ×2 stimuli at frequencies of 5 Hz or above.
Correlation of complex-spike activity with the vestibular stimulus during learning
The histograms in Figure 4 illustrate the relationship between the vestibular stimulus and the complex-spike activity of one HGVP during ×0 and ×2 stimuli at 0.5 and 5 Hz. Complex-spike activity precisely reflects the activity of the climbing-fiber input to the HGVP, because spikes in the climbing fiber produce complex spikes in the Purkinje cell target in a one-to-one manner. Complex-spike activity was modulated during all four stimuli and, at each frequency, satisfied the criterion that comparison with the vestibular stimulus discriminated the stimuli that decreased (×0) versus increased (×2) the gain of the VOR. For example, during the 5 Hz vestibular stimulus, complex-spike activity was in phase with ipsiversive head velocity when the gain of the VOR needed to decrease (Fig. 4B, ×0) but was in phase with contraversive head velocity when the gain needed to increase (Fig. 4D, ×2). However, the phase of the peak complex-spike response relative to the vestibular stimulus depended not just on whether the stimulus would induce a decrease or increase in the gain of the VOR but also on the stimulus frequency. During the ×0 stimuli, for example, complex-spike activity was in phase with ipsiversive head velocity at 5 Hz (Fig. 4B) but was in phase with contraversive head velocity at 0.5 Hz (Fig. 4A).
View larger version (18K):[in this window]
[in a new window]
Figure 4. Histograms showing climbing-fiber activity during stimuli that induce learned decreases (×0, A, B) and increases (×2, C, D) in the gain of the VOR. Climbing-fiber activity was recorded as complex spikes in the representative Purkinje cell whose simple-spike responses are shown in Figure 2. Climbing-fiber responses from 20-500 stimulus cycles were averaged to obtain each histogram. Note the different time scales in the left (A, C) and right (B, D) panels, which show data for sinusoidal stimuli at 0.5 and 5 Hz, respectively. Vertical dashed lines mark peak contraversive and ipsiversive head velocity. The simplified circuit diagram on the left highlights the loci of vestibular signals and climbing-fiber (CF) signals in the circuit for the VOR. IO, Inferior olive.
Figure 5 uses polar plots to summarize the correlation of the vestibular stimulus with the complex-spike responses in all 26 HGVPs in which the complex spike was isolated. For each stimulus frequency, the complex-spike responses to the ×0 and ×2 stimuli plot on opposite sides of the graph, forming distinguishable, almost nonoverlapping populations. However, in the different graphs, the phase of the complex-spike responses relative to the vestibular stimulus rotated as a function of stimulus frequency, indicating that the relative phase of climbing-fiber and vestibular signals was not similar for all stimuli that produced similar learned changes in the gain of the VOR. During the ×2 stimuli (Fig. 5, filled symbols), for example, peak complex-spike activity led ipsiversive head velocity slightly at 0.5 Hz (Fig. 5A), it lagged ipsiversive head velocity at 2 and 10 Hz (Fig. 5B,D), and it was almost exactly out of phase with ipsiversive head velocity at 5 Hz (Fig. 5C). Similarly, complex-spike responses to ×0 stimuli at different frequencies peaked during different phases of the vestibular stimulus (Fig. 5, + and X symbols).
View larger version (30K):[in this window]
[in a new window]
Figure 5. Summary of climbing-fiber responses, plotted relative to the vestibular stimulus. A-D, Climbing-fiber responses to stimulus frequencies of 0.5, 2, 5, and 10 Hz. Responses are plotted in polar coordinates, with distance from the origin corresponding to the amplitude of response modulation and an angle corresponding to the phase shift between peak climbing-fiber activity and head velocity. Responses in phase with peak ipsiversive head velocity are plotted to the right of the origin, responses in phase with peak contraversive head velocity are plotted to the left of the origin, and clockwise rotation around the graph represents increased phase lead. An individual climbing fiber contributed two symbols for each stimulus frequency: a + (monkey D) or X (monkey E) symbol for the ×0 stimulus, and a filled square (monkey D) or a filled triangle (monkey E) for the ×2 stimulus. R×0 marks the point in the vestibular stimulus leading peak contraversive head velocity by 46°, and R×2 marks the point in the vestibular stimulus leading peak ipsiversive head velocity by 46°. Open arrows show the predicted phase at each frequency for a fixed delay in the climbing-fiber response of 122 msec from R×0. Filled arrows show the predicted phase at each frequency for a fixed delay in the climbing-fiber response of 122 msec from R×2. In all panels, inner and outer circles represent 1 climbing-fiber spike per second (CFR/s) and 2 climbing-fiber spikes/sec, respectively. Note the difference in scale from Figure 3.
The change in phase of the complex-spike responses with frequency can be described by a fixed time delay between a fixed reference point in the vestibular stimulus and peak complex-spike activity. We fit the phases of the complex-spike responses to ×0 and ×2 stimuli at 0.5, 2, 5, and 10 Hz with the following equations:
(1) where
(2) ×0,
and
, T is a fixed time delay, R×0 is the reference point for ×0 stimuli, R×2 is the reference point for ×2 stimuli, and R×2 is constrained to be exactly 180° opposite R×0: R×2 = R×0 +
. We found that a fixed delay (T) of 122 msec and reference points on the vestibular stimulus leading peak contraversive (for R×0) and ipsiversive (for R×2) head velocity by 46° yielded the best fit. Thus, the frequency dependency of the complex spike responses is approximately consistent with the 100 msec latency between visual stimuli and complex-spike responses reported previously (Stone and Lisberger, 1990
The open and filled arrows in Figure 5 show the fixed time delay of 122 msec from R×0 and from R×2, respectively. The fixed delay translates into a progressively larger phase lag at higher frequencies (22°, 88°, 220°, and 439° at 0.5, 2, 5, and 10 Hz). Overall, the average peak complex-spike responses to ×0 stimuli at all four frequencies were well fit as occurring 122 msec after R×0 (Fig. 5, open arrows), and the average peak complex-spike responses to ×2 stimuli were well fit as occurring 122 msec after R×2 (Fig. 5, filled arrows). We constrained the reference points R×0 and R×2 to be 180° apart so that a single time delay would have the same meaning for both ×0 and ×2 stimuli. Although R×0 and R×2 were obtained by fitting the complex-spike responses, they correspond to points on the vestibular stimulus that coincide, within 30°, with peak activity of the primary afferents in the contralateral and ipsilateral vestibular nerves, respectively. On average, peak firing in the primary afferents leads peak ipsiversive head velocity by ~15° at 0.5 Hz and by ~55° at 8 Hz (Fernandez and Goldberg, 1971
Correlation of complex-spike and simple-spike activity during learning
Comparison of Figures 3 and 5 reveals that the correlation of simple-spike and complex-spike activity discriminates ×0 from ×2 stimuli during a subset, but not all of the stimulus frequencies tested. At 5 Hz, simple-spike and complex-spike activity were approximately in phase during the ×2 stimulus (Figs. 3C, 5C, filled symbols) and 180° out of phase during the ×0 stimulus (+ and X symbols). At 10 Hz (Figs. 3D, 5D), simple-spike and complex-spike activity were in phase during the ×0 stimulus and 180° out of phase during the ×2 stimulus. At 0.5 Hz (Figs. 3A, 5A) and 2 Hz (Figs. 3B, 5B), simple-spike and complex-spike activity were 180° out of phase during both ×0 and ×2 stimuli. Thus, the correlation of simple-spike and climbing-fiber signals discriminated stimuli that increase the gain of the VOR from stimuli that decrease the gain of the VOR at high stimulus frequencies, such as 5 or 10 Hz, but not at lower frequencies, such as 0.5 or 2 Hz. Even at the higher stimulus frequencies, the simple-spike and climbing-fiber signals failed the second criterion for signals that guide learning, of evincing similar patterns of activity during all stimuli that induce similar learned changes in the VOR.
Computational analysis of plasticity mechanisms driven by the correlation of climbing-fiber activity and vestibular signals
The only pair of signals examined that discriminated ×0 from ×2 stimuli at each stimulus frequency was the correlation of climbing-fiber (complex-spike) and vestibular signals. Therefore, this is the only pair of signals that potentially could guide learning across the full range of sinusoidal stimulus frequencies that induce learning. Moreover, the fits to Equations 1 and 2 suggest that a single plasticity mechanism must be tuned to climbing-fiber activity that follows activity in vestibular pathways by ~100 msec if it is to account for all motor learning in the VOR. To assess a number of factors that could affect this conclusion, we present a computational analysis of the features a plasticity mechanism must have to permit the correlation of climbing-fiber and vestibular signals to guide learning.
First, we must consider how the vestibular stimuli might be represented at the sites of plasticity. This issue is of particular importance in the cerebellar cortex, because the responses of vestibular parallel fibers have not been described, and the question of how to identify granule cell recordings in behaving animals has not been resolved. In vestibular primary afferents, sinusoidal vestibular stimuli are represented by sinusoidal modulation of firing rate. At frequencies of 0.5, 2, 5, and 10 Hz, firing rate leads ipsiversive head velocity by phases that average ~15°, 25°, 40°, and 60° (Fernandez and Goldberg, 1971
We demonstrate that a plasticity mechanism tuned to a 100 msec delay between its vestibular and climbing-fiber inputs is required to account for motor learning in the VOR, even if the vestibular inputs do exhibit a broad range of response phases. We further show that this tuning must be implemented in the plasticity mechanism itself and not by using neural circuits to delay the incoming vestibular signals.
Neural circuit model
In our simulations, we consider a single Purkinje cell with 36 vestibular parallel-fiber inputs, five of which are illustrated in Figure 6. We assume that each parallel fiber exhibits sinusoidal modulation of its firing rate during a sinusoidal head velocity stimulus. Each parallel fiber has a unique phase of peak activity, distributed evenly from
90° to +260° with respect to head velocity:
(3) where t is time, and
) correspond to parallel fibers with progressively more phase lag. Our simulations thus consider the full range of possible response phases in the vestibular parallel fibers.
View larger version (16K):[in this window]
[in a new window]
Figure 6. Schematic showing several simulated vestibular parallel-fiber (PF) inputs to a single Purkinje cell (PC). The trace above each parallel fiber shows its activity during sinusoidal head rotation about a vertical axis. The dashed vertical line marks the time of peak ipsiversive head velocity. Activity in PF lags peak ipsiversive head velocity by
Equations 4-7 below are a formal description of widely accepted principles about functional connectivity in the circuit for the VOR. We used these equations to calculate the predicted changes in the VOR resulting from changes in the weights of the vestibular inputs to the Purkinje cell from parallel fibers with different phases. We assume that the VOR-driven eye movement response before learning is equal in amplitude and opposite in direction to head velocity (H). Thus, during a sinusoidal vestibular stimulus of amplitude A, the prelearning VOR-driven eye movement is described by:
Positive values for H(t) and VORpre(t) represent ipsiversive head or eye velocities. The VOR-driven eye movement response after learning (VORpost) is equal to the sum of the prelearning VOR (VORpre) and the change in the VOR attributable to learning (
(4) VOR):
Activation of Purkinje cells in the floccular complex produces ipsiversive eye movement by inhibiting neurons in the direct VOR pathways through the brainstem (Baker et al., 1972
(5) where K is a constant >0 (K = 1.0 in all simulations).
(6) We can describe the learned change in activity of the Purkinje cell attributable to learned changes in the parallel-fiber weights as:
where w
(7) A critical feature of the circuit for the VOR captured in Equations 3-7 is that the timing of activity in a Purkinje cell will determine the effect of that activity on the gain of the VOR. Purkinje cell activity drives ipsiversive eye movement (or, equivalently, it reduces contraversive eye movement). Enhanced ipsiversive eye movement during contraversive head movement constitutes an increase in the gain of the VOR. Enhanced ipsiversive eye movement (reduced contraversive eye movement) during ipsiversive head movement constitutes a decrease in the gain of the VOR, and thus:
Increasing Purkinje cell activity during contraversive head movement increases the gain of the VOR.
Increasing Purkinje cell activity during ipsiversive head movement decreases the gain of the VOR.
The timing of Purkinje cell activity relative to head movement is determined by the balance of synaptic input from parallel fibers that fire at different phases of the head movement. Thus, if multiple parallel-fiber weights are changed, the effect on the VOR will depend on the change in the balance of synaptic strength in vestibular parallel fibers with different phases. A relative increase in the synaptic strength of parallel fibers that fire most during contraversive head velocity will increase the gain of the VOR. A relative increase in the synaptic strength of parallel fibers that fire most during ipsiversive head velocity will decrease the gain of the VOR. We will consider primarily the effects of synaptic depression rather than potentiation and the corresponding principles:
If depression is greater in the parallel fibers that fire most during contraversive head velocity, then the gain of the VOR decreases.
If depression is greater in the parallel fibers that fire most during ipsiversive head velocity, then the gain of the VOR increases.
Figures 7 and 8 illustrate the above principles by showing the predicted effects on the VOR for changes in the weights of individual parallel fibers with different phases. Figure 7 illustrates the values of the variables in Equations 3-7 during two cycles of a sinusoidal vestibular stimulus (H). In each panel, the synaptic weight of one of the 36 parallel fibers was decreased: PF0 in A, PF90 in B, and PF180 in C. In Figure 7, the bottom three traces show the predicted effects of changing the weight of that single parallel fiber on Purkinje cell activity during the vestibular stimulus (
View larger version (13K):[in this window]
[in a new window]
Figure 7. Predicted learned changes in the VOR for a reduction in the weight of parallel fibers that fire at different phases of the vestibular stimulus. Traces representing the variables in Equations 3-7 are shown for a reduction in the weight of a parallel fiber whose activity peaks during ipsiversive head velocity (A,
View larger version (16K):[in this window]
[in a new window]
Figure 8. Predicted learned changes in the gain (A) and phase (B) of the VOR, plotted as a function of the phase of the vestibular parallel fiber (PF) undergoing synaptic depression (LTD) (Eq. 7,
Figure 8 summarizes the changes in the gain (A) and the phase (B) of the VOR that are predicted for reductions in the weights of individual parallel fibers whose activity peaks at different phases of the vestibular stimulus. Changes in the gain and phase of the VOR were obtained by first computing VORpost(t) and measuring its amplitude and phase. The gain of the VOR after learning was then computed as the amplitude of VORpost(t) divided by the amplitude of H(t). This postlearning gain value was also equal to the change in VOR gain (post/pre), because the gain of the VOR before learning was assumed to be 1.0 (Eq. 4). Note that computation of VORpost(t) involves summation of multiple sine waves (Eqs. 3-7) and that the amplitude of VORpost(t) depends on both the amplitude and phase of those sine waves. In Equation 5, for example, the amplitude of VORpost(t) is not equal to the sum of the amplitude of VORpre(t) and the amplitude of
For each parallel fiber phase
Simultaneous plasticity mechanism
Figures 7 and 8 illustrate that to understand how plasticity mechanisms with different features will affect the gain of the VOR, we need to calculate how those plasticity mechanisms will change the relative weights of vestibular parallel fibers that exhibit peak activity during different phases of the vestibular stimulus. The first plasticity mechanism evaluated was one that decreased the weight of a parallel fiber in proportion to the level of activity in that parallel fiber at the time of a climbing-fiber spike:
where T1, T2, ... , Tj, ... , Tk are the times of spikes in the climbing fiber, and B is a constant >0. To avoid possible artifacts that might result from using any type of fit to approximate the climbing-fiber responses, the values of Tj were obtained from the actual complex-spike times, recorded to a resolution of 1 msec, during a 60-120 sec epoch of a ×0 or ×2 stimulus. Each Tj was used once in a simulation, and division by k normalized for the number of climbing-fiber spikes used in each simulation. Multiplication by
![&Dgr;w<SUB>ϕ</SUB>=<LIM><OP>∑</OP></LIM><SUB>(j<UP>=1:</UP>k)</SUB>[<UP>−</UP>B PF<SUB>ϕ</SUB>(T<SUB>j</SUB>)]/k,](/content/vol18/issue21/fulltext/9112/img009.gif)
(8) Figure 9 summarizes the results of simulations that used Tj values obtained from the complex spike responses shown in Figure 4 to drive the simultaneous plasticity mechanism. The two graphs on the left side of Figure 9, A1 and A2, show the predicted changes in the weights of the different vestibular parallel fibers for 0.5 and 5 Hz stimuli. These two graphs are summarized in Figure 9B, which plots the phase of the parallel fiber with the greatest synaptic depression as a function of the frequency of the ×0 and ×2 stimuli. Finally, the predicted changes in all parallel-fiber weights were converted to predicted changes in the VOR using Equations 3-7, and the predicted changes in the gain of the VOR are plotted as a function of stimulus frequency in Figure 9C.
View larger version (23K):[in this window]
[in a new window]
Figure 9. Predicted synaptic and behavioral changes produced by a plasticity mechanism driven by simultaneous activity in climbing fibers and vestibular parallel fibers. Spike trains from the typical climbing fiber, whose responses are shown in Figure 4, were used as the input to the simultaneous plasticity mechanism (Eq. 8). Open symbols, Predicted changes for the climbing fiber spike trains present during ×0 stimuli; filled symbols, predicted changes for the climbing fiber spike trains present during ×2 stimuli. A, Predicted changes in the synaptic weights of parallel fibers that fire at different phases of the vestibular stimulus. A1, Changes predicted for 0.5 Hz stimuli. A2, Changes predicted for 5 Hz stimuli. B, Phase of the parallel fiber undergoing largest weight reduction (LTD) as a function of the stimulus frequency. The thin vertical lines to the right of the graph mark the range of phases for ×0 stimuli at frequencies of 0.5, 2, 5, and 10 Hz, and the thick vertical lines represent the range of phases for ×2 stimuli. C, Predicted change in the gain of the VOR as a function of stimulus frequency. Arrows of the same style mark results corresponding to the same simulation.
During a 0.5 Hz, ×0 stimulus, most Tj values (complex spike times) coincided with contraversive head motion (Fig. 4A), so the greatest weight changes occurred in parallel fibers that fired most vigorously during contraversive head motion (Fig. 9A1,B, open single arrows). As a result, the predicted change in the VOR was an appropriate decrease in gain (Fig. 9C, open single arrow). During the 0.5 Hz, ×2 stimulus, most Tj values coincided with ipsiversive head motion (Fig. 4C), the biggest weight changes occurred in parallel fibers that fired most vigorously during ipsiversive head motion (Fig. 9A1,B, filled single arrows), and the predicted change in the VOR was an appropriate increase in gain (Fig. 9C, filled single arrow). Thus, as suggested previously (Ito, 1972
In contrast, the simultaneous plasticity mechanism fails to predict the induction of appropriate changes in the VOR by the activity of this climbing fiber during the 5 Hz stimuli. During the 5 Hz, ×2 stimulus, most Tj values coincided with contraversive head motion (Fig. 4D), so the parallel fibers that fired during contraversive head motion underwent the most weight reduction (Fig. 9A2,B, filled double arrows). These changes in synaptic weight predicted a decrease in the gain of the VOR (Fig. 9C, filled double arrow), which is opposite to the gain change observed experimentally. Similarly, the simultaneous plasticity mechanism incorrectly predicted an increase in the gain of the VOR when the complex spikes recorded during the 5 Hz, ×0 stimulus (Fig. 4B) provided Tj values (Fig. 9C, double open arrow). Similar computations were performed using Tj values obtained from the complex-spike times recorded in the same cell during 2 and 10 Hz stimuli, and the results are plotted in Figure 9, B and C. For these stimulus frequencies, predicted changes in gain were in the correct direction.
Figure 9B illustrates that the parallel fiber undergoing the biggest weight change varied with the frequency of the stimulus, as well as the stimulus configuration (×0 vs ×2). For the climbing fiber in this example, the parallel fibers undergoing the most depression during ×2 stimuli at frequencies from 0.5 to 10 Hz spanned a range of phases from
We have focused on the parallel-fiber weights that underwent the greatest decrease under each stimulus condition. Inspection of Figure 9, A1 and A2, reveals that there were decreases in the weights of all parallel fibers. This results from our use of a plasticity mechanism that yields exclusively synaptic depression. We obtained the same effects on the gain of the VOR with a mixture of increases and decreases in parallel-fiber weights if we used a plasticity mechanism like that suggested by Bienenstock et al. (1982)
We performed similar computations to those illustrated in Figure 9 using Tj values derived from the eight other individual HGVPs in which complex-spike times were recorded during all four stimulus frequencies. The results shown in Figure 9 are typical. For none of the nine cells did computations using the simultaneous plasticity mechanism predict appropriate gain changes for all stimulus frequencies. Thus, a plasticity mechanism guided by coincident activity in climbing fibers and vestibular inputs cannot account for the induction of appropriate learned changes in the gain of the VOR across the range of effective stimulus frequencies.
Nonsimultaneous plasticity mechanism
We next evaluated whether a plasticity mechanism tuned to nonsimultaneous activity in climbing fibers and vestibular parallel fibers might provide consistent guidance for learning across the range of effective stimulus frequencies:
where TPF-CF is a constant time delay we will refer to as the parallel fiber to climbing fiber (PF-CF) interval. When TPF-CF = 0, the nonsimultaneous plasticity mechanism described by Equation 9 is equivalent to the simultaneous plasticity mechanism of Equation 8. When TPF-CF
![&Dgr;w<SUB>ϕ</SUB>=<LIM><OP>∑</OP></LIM><SUB>(j=1:k)</SUB>[<UP>−</UP>B PF<SUB>ϕ</SUB>(T<SUB>j</SUB>−T<SUB>PF–CF</SUB>)]/k,](/content/vol18/issue21/fulltext/9112/img010.gif)
(9) 0, climbing-fiber spikes induce a change in the weight of each parallel fiber proportional to the level of activity in that parallel fiber TPF-CF milliseconds before (for positive TPF-CF) or after (for negative TPF-CF) the climbing-fiber spike.
Figure 10 compares the results for several different values of TPF-CF when complex spike times from the example in Figure 4 provided the Tj values used by the nonsimultaneous plasticity mechanism. Changes in the weights of each of the 36 parallel fibers were computed for the ×0 and ×2 stimuli from Equation 9, with TPF-CF = 50 msec (Fig. 10A,D), TPF-CF = 100 msec (Fig. 10B,E), or TPF-CF = 200 msec (Fig. 10C,F). As in Figure 9B, the top panels of Figure 10 plot the phase of the vestibular parallel fiber that underwent the largest reduction in weight as a function of the stimulus frequency. The predicted changes in all the parallel-fiber weights were used to compute predicted changes in the gain of the VOR from Equations 3-7 and were plotted as a function of the stimulus frequency in the bottom panels of Figure 10. When TPF-CF was 100 msec, parallel fibers with a phase close to 180° (those that fire most during contraversive head velocity) underwent the greatest weight reduction for ×0 stimuli at all frequencies tested, and parallel fibers with a phase of ~0° (those that fire most during ipsiversive head velocity) underwent the greatest weight reduction for all ×2 stimuli (Fig. 10B). There was no overlap in the sets of parallel fibers undergoing the greatest weight reduction in response to the ×0 versus ×2 stimuli (thin and thick vertical lines at right edge of graph). These changes in parallel-fiber weights for TPF-CF = 100 msec predicted consistent decreases in the gain of the VOR in response to the ×0 stimuli and consistent increases in the gain of the VOR in response to the ×2 stimuli for the full range of stimulus frequencies included in the experiments (Fig. 10E). In contrast, when TPF-CF = 50 or 200 msec, there was considerable variation in the phase of the parallel fibers undergoing the greatest weight changes for ×0 or ×2 stimuli at different frequencies, and there was overlap in the range of parallel fibers undergoing the largest weight reductions in response to ×0 versus ×2 stimuli (Fig. 10A,C). These changes in parallel-fiber weights predicted appropriate gain changes at some frequencies and inappropriate gain changes at other frequencies (Fig. 10D,F).
View larger version (19K):[in this window]
[in a new window]
Figure 10. Predicted synaptic and behavioral changes for a plasticity mechanism driven by nonsimultaneous activity in climbing fibers and vestibular parallel fibers. Spike trains from the typical climbing fiber in Figure 4 were used as the input to the nonsimultaneous plasticity mechanism (Eq. 9). Reduction in the parallel-fiber weights was proportional to the activity in the parallel fiber at some interval (TPF-CF) before a spike in the climbing fiber. Results are shown for three different values of TPF-CF: 50, 100, and 200 msec. A-C, Phase of the parallel fiber undergoing largest synaptic weight reduction as a function of the stimulus frequency. The thin vertical lines to the right of each graph mark the range of phases for ×0 stimuli at frequencies of 0.5, 2, 5, and 10 Hz, and the thick vertical lines represent the range of phases for the ×2 stimuli. D-F, Predicted change in the gain of the VOR as a function of stimulus frequency. Open symbols, Predicted changes for the climbing-fiber spike trains present during ×0 stimuli; filled symbols, predicted changes for the climbing-fiber spike trains present during ×2 stimuli.
Figure 11 delimits the range of values for the PF-CF interval (TPF-CF) that predicted appropriate changes in the gain of the VOR using the complex-spike times from the example in Figure 4 as the input Tj values. These ranges were obtained from computations performed using values of TPF-CF from
View larger version (44K):[in this window]
[in a new window]
Figure 11. Range of PF-CF intervals (range of values for TPF-CF) yielding appropriate changes in the VOR. Spike trains from the typical climbing fiber in Figure 4 were used as the input to the nonsimultaneous plasticity mechanism (Eq. 9). Predicted changes in the gain of the VOR for ×0 stimuli (A-D) and ×2 stimuli (E-H) are plotted as a function of the value of TPF-CF. A, E, Predictions for 0.5 Hz stimuli. B, F, Predictions for 2 Hz stimuli (bold traces) and 0.5 Hz stimuli (thin traces). C, G and D, H, Predictions for 5 and 10 Hz stimuli (bold traces), with predictions for lower frequencies replotted (thin traces). Shaded regions in each graph mark the range of values for TPF-CF that yielded appropriate predicted changes in the gain of the VOR for all stimuli at or below the indicated frequency.
We define as "effective" any PF-CF interval that predicts decreases in the gain of the VOR for all ×0 stimuli and increases in gain for all ×2 stimuli. Any PF-CF interval with a trace plotting above 1.0 for a ×0 stimulus or below 1.0 for a ×2 stimulus cannot be considered effective, because that PF-CF interval fails to satisfy the criterion of predicting a change in the gain of the VOR in the experimentally observed direction across all stimulus frequencies. The shaded regions in each panel of Figure 11 mark successive approximations of the range of effective PF-CF intervals as additional stimulus frequencies are considered. These shaded regions show that the inclusion of more and higher stimulus frequencies increasingly narrows the range of PF-CF intervals that could be effective. For ×0 stimuli, values of TPF-CF from ~90 to 130 msec predicted decreases in the gain of the VOR at all stimulus frequencies tested (Fig. 11D). For ×2 stimuli, values of TPF-CF from ~80 to 120 msec predicted increases in the gain of the VOR at all stimulus frequencies tested (Fig. 11H). Other values of TPF-CF predicted inappropriate changes in the gain of the VOR at one or more stimulus frequency. The computations predicted only small changes in the phase of the VOR, <20° for all values of TPF-CF and all stimulus frequencies (not plotted).
Figure 12 summarizes the effective PF-CF intervals estimated for the climbing-fiber inputs to each of the nine HGVPs in which complex-spike responses to all four stimulus frequencies were recorded. Each bold horizontal line in Figure 12 shows the range of values of TPF-CF that yielded appropriate changes in the gain of the VOR at all four frequencies when a different climbing fiber provided the input Tj values to the nonsimultaneous plasticity mechanism. For all nine climbing fibers, a range of values of TPF-CF around 100 msec predicted appropriate changes in gain for both ×0 and ×2 stimuli. For five cells, a TPF-CF value of approximately
View larger version (15K):[in this window]
[in a new window]
Figure 12. Range of PF-CF intervals (range of values for TPF-CF) that yielded appropriate predicted changes in the gain of the VOR for all four stimulus frequencies tested (0.5, 2, 5, and 10 Hz). Each line represents the range of effective values of TPF-CF for one individual climbing fiber. A, Range of values of TPF-CF that yielded decreases in gain in response to ×0 stimuli. B, Range of values of TPF-CF that yielded increases in gain in response to ×2 stimuli.
Figure 12 shows unequivocally that a value for TPF-CF of 0 msec failed to predict appropriate changes in VOR gain for even a single climbing fiber. Therefore, plasticity mechanisms tuned to coincident activity in climbing-fiber and vestibular parallel-fiber pathways cannot provide consistent guidance for learning in the VOR. To account for learning under all stimulus conditions tested, a plasticity mechanism should be specifically "tuned" to climbing-fiber activity that follows activity in the vestibular inputs by ~100 msec. The small difference between this 100 msec estimate of the time across which vestibular and climbing-fiber inputs should be compared and the value of 122 msec obtained from the fits to Equations 1 and 2 may be attributed to the use of the actual time of each complex spike for the simulations versus a sinusoidal fit of the complex-spike responses in Equations 1 and 2.
Could the simultaneous plasticity mechanism be used if circuit properties delayed the vestibular parallel fiber inputs to Purkinje cells?
One potential implementation of the nonsimultaneous plasticity mechanism would be to use neural circuit properties to delay the vestibular signals in the parallel fibers and then use a simultaneous plasticity mechanism to change synaptic weights. However, Figure 13 illustrates that this mechanism would introduce delays in the learned, parallel-fiber-mediated eye movements that would result in incorrect changes in the VOR. The effects of delaying the vestibular signals in the parallel fibers on the predicted changes in the VOR are shown in Figure 13 for ×2 stimuli at 0.5 Hz (left) and 5 Hz (right). For the purpose of illustration, we used only two vestibular afferents in this simulation, one whose activity peaked during peak ipsiversive head velocity and one whose activity peaked during peak contraversive head velocity (Fig. 13, AFFi, AFFc). We introduced a 100 msec delay between activity in the primary vestibular afferents driven by head movement and activity in the vestibular parallel fibers driven by the primary afferents (Fig. 13, PFi, PFc). This 100 msec delay translates into a phase lag of 18° at 0.5 Hz and 180° at 5 Hz, causing elevated activity in PFi to coincide with elevated climbing fiber activity (CFR) during ×2 stimuli at both frequencies. Therefore, a simultaneous LTD plasticity mechanism would produce the most synaptic depression in the PFi during ×2 stimuli at both frequencies. The weights of the two parallel fibers (wi, wc) were initially set to an equal and arbitrary value. The ×2 stimuli at both frequencies induced more depression in wi than in wc. This frequency-independent change in the balance of input to the Purkinje cell from parallel fibers driven by ipsiversive and contraversive head movement can be seen by comparing PFi × wi and PFc × wc before (Fig. 13, dashed trace) and after (Fig. 13, solid trace) learning at both frequencies.
View larger version (24K):[in this window]
[in a new window]
Figure 13. Predicted synaptic and behavioral changes for a simultaneous plasticity mechanism when a 100 msec delay was introduced in the vestibular parallel-fiber signals. Predictions were calculated for ×2 stimuli at 0.5 and 5 Hz, using two parallel fibers whose vestibular signals lagged ipsiversive or contraversive head velocity by 100 msec. H, Head velocity; AFFi, AFFc, activity in vestibular afferents that fire in phase with ipsiversive and contraversive head velocity, respectively; PFi, PFc, activity in parallel fibers whose responses are delayed 100 msec relative to those in AFFi, AFFc (note that the different time scales in the left and right panels affect the apparent size of the 100 msec shift); CFR, average responses in the climbing fiber shown in Figure 4; PFi × wi, PFc × wc, input to the Purkinje cell from each of the vestibular parallel fibers before (dashed trace) and after (solid trace) learning, computed as activity in the parallel fiber multiplied by its synaptic weight;
We then used Equations 4-7 to determine how the reduced weights of the delayed vestibular parallel-fiber signals would affect the gain of the VOR. Comparison of VORpost (Fig. 13, solid trace) and VORpre (dashed trace) shows a correct prediction of a learned increase in the gain of the VOR for the ×2 stimulus at 0.5 Hz and an incorrect prediction of a learned decrease in the gain of the VOR for the ×2 stimulus at 5 Hz. Delaying the vestibular parallel fiber inputs to Purkinje cells delays the effect of those vestibular inputs on eye movements so that they are 180° late at 5 Hz. Elevated activity in PFi coincided with ipsiversive head movement at 0.5 Hz, but the 100 msec delay caused elevated activity in PFi to coincide with contraversive head movement at 5 Hz. Thus, reduction in the synaptic weight of PFi had opposite effects at the two frequencies, causing an increase in the gain of the VOR at 0.5 Hz and a decrease at 5 Hz. This example illustrates that filtering or delaying the vestibular signals by network properties cannot enable a plasticity mechanism tuned to simultaneous climbing-fiber and vestibular inputs to provide consistent guidance for motor learning in the VOR.
Effect of sharpness of tuning in the nonsimultaneous plasticity mechanism
Finally, we analyze how learning would be affected by the sharpness of tuning to the optimal TPF-CF of ~100 msec. In the above computations, plasticity was very sharply tuned: the amount of plasticity induced in a parallel fiber by a spike in a climbing fiber was solely determined by the activity in that parallel fiber exactly TPF-CF milliseconds before the climbing-fiber spike. Here, we modify the plasticity mechanism so that the amount of plasticity induced by a spike in a climbing fiber depends on parallel-fiber activity in a window of time surrounding TPF-CF milliseconds before the climbing-fiber spike. This is accomplished by convolving parallel-fiber activity with a Gaussian g(t), centered TPF-CF milliseconds before the climbing-fiber spike:
![&Dgr;w<SUB>ϕ</SUB>=<LIM><OP>∑</OP></LIM><SUB>(j=1:k)</SUB> (<UP>−</UP>B/k)[PF<SUB>ϕ</SUB> ∗ g(T<SUB>j</SUB>)],](/content/vol18/issue21/fulltext/9112/img011.gif)
(10)
with B = 4 × 10
4. The * represents convolution, which can be thought of as a second-order process (e.g., the concentration of a second messenger), rising and decaying in response to parallel-fiber activity.
The value of
determines the sharpness of tuning of the plasticity mechanism. Figure 14 shows the predicted changes in the gain of the VOR as a function of
View larger version (37K):[in this window]
[in a new window]
Figure 14. Predicted change in the gain of the VOR as a function of the width of tuning of the plasticity mechanism. Predictions for stimulus frequencies of 0.5, 2, 5, and 10 Hz are shown in separate plots. Responses of nine individual climbing fibers were used as the input to the nonsimultaneous plasticity mechanism (Eq. 10), with a fixed value for TPF-CF of 100 msec. Reduction in the weight of a parallel fiber was proportional to activity in that parallel fiber in an eligibility window centered 100 msec before a spike in the climbing fiber. Changes in the VOR predicted from the changes in parallel fiber weights are shown for eligibility windows of various widths, defined by values for
Behavioral studies have shown that ×0 and ×2 stimuli at a frequency of 5 Hz are approximately comparable in their effectiveness in inducing learning in the VOR as ×0 and ×2 stimuli at 2 Hz but that 10 Hz stimuli are considerably less effective at inducing learning than 5 Hz stimuli (Raymond and Lisberger, 1996
Insensitivity of predictions to features of the model
Our recording results indicate that a plasticity mechanism driven by simultaneous activity in climbing-fiber and vestibular parallel-fiber pathways cannot account for motor learning in the VOR. The simulation results in Figures 6-12 illustrate this point for the specific case in which vestibular parallel fibers exhibit sinusoidal modulation in their firing and a broad range of response phases, but the conclusion is not dependent on the particular form of the vestibular representation at the site of plasticity. Because Purkinje cell activity drives ipsiversive eye movement, a decrease in the amplitude of the VOR would require a decrease in the excitatory drive to Purkinje cells during contraversive head movement. This means decreasing the synaptic strength of whichever parallel fibers fire most vigorously during contraversive head movement, regardless of their specific firing rate profiles or how those responses are driven by various head movement parameters. During ×0 stimuli at 5 Hz, LTD driven by simultaneous climbing-fiber and parallel-fiber activity could not depress the parallel fibers active during contraversive head movement, because the climbing fibers were not active during contraversive head movement.
We confirmed that our fundamental conclusions were valid for other types of vestibular representation by running simulations in which the vestibular parallel fibers covered a more limited range of phases and clustered in two populations whose activity peaked near peak ipsiversive head velocity (V-i) and peak contraversive head velocity (V-c), and simulations in which the parallel fibers exhibited bursts of activity at different times in the head movement rather than sinusoidally modulated firing rates. Results for these simulations were similar to those in Figures 6-12.
We also want to emphasize that our analysis is not affected by the controversy that has existed about the sites of plasticity for motor learning in the VOR (Ito, 1993
DISCUSSION
Previously, two mechanisms of motor learning in the VOR had been hypothesized. The first suggested that simultaneous activity in visual climbing fibers and vestibular parallel fibers would induce LTD in the cerebellar cortex that could mediate motor learning (Ito, 1972
Is motor learning in the VOR mediated by a single mechanism driven by nonsimultaneous climbing-fiber and vestibular activity?
The climbing fiber recordings at high-stimulus frequencies (e.g., 5 Hz) reveal that a plasticity mechanism driven by simultaneous activity in climbing-fiber and vestibular parallel-fiber pathways cannot account for motor learning in the VOR. The simulations further demonstrate that a comparison of climbing-fiber spikes with the vestibular signals present ~100 msec earlier contains the information required to guide learning. The question of how to implement a neural learning rule in the brain that compares events across time represents a general challenge in linking cellular and systems-level analyses of associative learning. Cellular studies have emphasized the importance of temporal coincidence, of either presynaptic and postsynaptic activity or of activity in two presynaptic inputs, in the induction of associative plasticity. The assumption has been that plasticity mechanisms that detect coincident neural activity would be useful for detecting events that are paired in time during associative learning. However, at the behavioral level, associative learning often involves the detection and comparison of events that are related in time but nonsimultaneous. For example, classical conditioning is most effectively induced when an unconditioned stimulus (e.g., an electric shock) follows a conditioned stimulus (e.g., a tone) by some nonzero interval.
Learning in the VOR is analogous to classical conditioning in that the pairing of two stimuli (head turns and visual image motion) results in learned changes in the eye movement response to one of the stimuli (head turns) (Raymond et al., 1996
Models of classical conditioning have adopted two basic strategies for bridging the time interval between nonsimultaneous events. The first hypothesizes that circuit properties bridge the gap by delaying or filtering neural signals related to one of the events. During cerebellum-dependent classical conditioning of the eyelid response, for example, mossy-fiber activity related to the conditioned stimulus (tone) precedes climbing-fiber activity related to the unconditioned stimulus (shock). It has been suggested that granule-Golgi pathways in the cerebellum could delay and filter the mossy fiber signals so that the resultant parallel-fiber activity would coincide with the climbing-fiber activity (Fujita, 1982
One might expect that for the VOR, a nonzero TPF-CF in Equations 9 and 10 could be implemented by either mechanism. However, we found that introducing a delay in the vestibular signals carried by parallel fibers did not work, because it delayed both the signals that guide plasticity and their contribution to motor output. In contrast, if the delayed representation of a stimulus used for the induction of plasticity is a biochemical eligibility trace produced by nondelayed parallel-fiber activity, then the timing of the learned behavior need not be delayed. This latter property is required to account for learning in the VOR induced by high-frequency stimuli.
The suggestion of an eligibility trace with a 100 msec delay is difficult to reconcile with most cellular studies, which have used electrical or pharmacological stimulation to study LTD in the cerebellar cortex. These generally have found that LTD is most effectively induced when climbing-fiber activity is simultaneous with or precedes parallel-fiber stimulation (Ito et al., 1982
Our data do not necessitate the abandonment of cerebellar LTD as a putative cellular mechanism of learning in the VOR. However, our analysis highlights the importance of new experiments to determine whether LTD in the functioning cerebellum has temporal contingencies suitable for guiding learning in response to the natural patterns of input activity that are actually present during learning. Further experiments should also examine the possibility that plasticity in the vestibular nucleus might be regulated by climbing-fiber and vestibular inputs with the required temporal precision.
Is motor learning in the VOR mediated by multiple mechanisms driven by different combinations of neural signals?
If plasticity in the functioning cerebellum is tuned to simultaneous climbing-fiber and parallel-fiber activity, then there must exist at least two mechanisms for motor learning in the VOR that contribute differentially to learning induced by high- and low-frequency stimuli. The results in Figure 3 indicate that the correlation of vestibular and Purkinje cell simple-spike signals cannot account for the induction of learning by high-frequency stimuli but could contribute to learning during low-frequency stimuli. Likewise, the correlation of simple-spike and climbing-fiber signals cannot account for all learning but could make a contribution during high-frequency stimuli.
Several previous observations are consistent with the idea of multiple mechanisms of learning in the VOR. First, behavioral studies have suggested the existence of two components of learning in the VOR that are differentially induced by high- and low-frequency stimuli (Raymond and Lisberger, 1996
Together, these results raise the possibility that the mechanisms underlying learning during the low- and high-frequency stimuli may be different. Learning could be guided by the correlation of the vestibular input with either climbing-fiber activity or simple spikes in the Purkinje cells during low-frequency stimuli, and it could be guided by the correlation of climbing-fiber activity with either the vestibular input or Purkinje cell simple spikes during high-frequency stimuli. We are not suggesting that the brain analyzes the frequency of the vestibular stimulus and selects from a menu of available mechanisms. Rather, there may be multiple mechanisms that contribute differentially depending on the available neural signals. The use of multiple cellular mechanisms of plasticity and error signals may ensure that learning can be induced by stimuli with a wide range of temporal properties.
FOOTNOTES Received Feb. 5, 1998; revised Aug. 17, 1998; accepted Aug. 20, 1998.
This work was supported by National Institutes of Health Grants EY10198, EY03878 (S.G.L.), and DC03342 (J.L.R.), and a National Aeronautics and Space Administration Research Associate Fellowship (J.L.R.). We thank members of the Lisberger and Sejnowski laboratories and D. Buonomano, H. Mahncke, and R. Venturini for many helpful comments on earlier drafts of this manuscript. We especially thank M. Kahlon for input throughout the writing process.
Correspondence should be addressed to Jennifer L. Raymond, Department of Physiology, Box 0444, University of California at San Francisco, San Francisco, CA 94143.
REFERENCES
- Albus JS (1971) A theory of cerebellar function. Math Biosci 10:25-61.
- Baker R, Precht W, Llinas R (1972) Cerebellar modulatory action on the vestibulo-trochlear pathway in the cat. Exp Brain Res 15:364-385[Web of Science][Medline].
- Balaban CD, Kawaguchi Y, Watanabe E (1981) Evidence of a collateralized projection from the inferior olive to the flocculus and vestibular nuclei in rabbits. Neuroscience Lett 22:23-29[Medline].
- Bienenstock EL, Cooper LN, Munro PW (1982) Theory for the development of neuron selectivity: orientation specificity and binocular interaction in visual cortex. J Neurosci 2:32-48[Abstract].
- Bock O (1982) Non-linear interaction of the vestibular and the eye tracking system in man. Exp Brain Res 47:461-464[Medline].
- Bullock D, Fiala JC, Grossberg S (1994) A neural model of timed response learning in the cerebellum. Neural Networks 7:1101-1114.
- Buonomano DV, Mauk MD (1994) Neural network model of the cerebellum: temporal discrimination and the timing of motor responses. Neural Comput 6:38-55[Web of Science].
- Chapeau-Blondeau F, Chauvet G (1991) A neural network model of the cerebellar cortex performing dynamic associations. Biol Cybern 65:267-279[Web of Science][Medline].
- Chen C, Thompson RF (1995) Temporal specificity of long-term depression in parallel fiber-Purkinje synapses in rat cerebellar slice. Learn Mem 2:185-198
[Abstract/Free Full Text] .- Dufosse M, Ito M, Jastreboff PJ, Miyashita Y (1978) A neural correlate in rabbit's cerebellum to adaptive modification of the vestibulo-ocular reflex. Brain Res 150:611-616[Web of Science][Medline].
- duLac S, Raymond JL, Sejnowski TJ, Lisberger SG (1995) Learning and memory in the vestibulo-ocular reflex. Annu Rev Neurosci 18:409-441[Web of Science][Medline].
- Ekerot C-F, Kano M (1985) Long-term depression of parallel fibre synapses following stimulation of climbing fibres. Brain Res 342:357-360[Web of Science][Medline].
- Ekerot C-F, Kano M (1989) Stimulation parameters influencing climbing fibre induced long-term depression of parallel fibre synapses. Neurosci Res 6:264-268[Web of Science][Medline].
- Fernandez C, Goldberg JM (1971) Physiology of peripheral neurons innervating semicircular canals of the squirrel monkey. II. Response to sinusoidal stimulation and dynamics of peripheral vestibular system. J Neurophysiol 34:661-675
[Free Full Text] .- Fiala JC, Grossberg S, Bullock D (1996) Metabotropic glutamate receptor activation in cerebellar Purkinje cells as substrate for adaptive timing of the classically conditioned eye-blink response. J Neurosci 16:3760-3774
[Abstract/Free Full Text] .- Fuchs AF (1967) Periodic eye tracking in the monkey. J Physiol (Lond) 193:161-171[Medline].
- Fujita M (1982) Adaptive filter model of the cerebellum. Biol Cybern 45:195-206[Web of Science][Medline].
- Fukuda J, Highstein SM, Ito M (1972) Cerebellar inhibitory control of the vestibulo-ocular reflex investigated in rabbit 3rd nucleus. Exp Brain Res 14:511-526[Web of Science][Medline].
- Gauthier GM, Robinson DA (1975) Adaptation of the human vestibulo-ocular reflex to magnifying lenses. Brain Res 92:331-335[Web of Science][Medline].
- Gerrits NM, Voogd J (1989) The topographical organization of climbing and mossy fiber afferents in the flocculus and ventral paraflocculus in rabbit, cat, and monkey. Exp Brain Res Suppl 17:26-29.
- Goldreich D, Krauzlis RJ, Lisberger SG (1992) Effect of changing feedback delay on spontaneous oscillations in smooth pursuit eye movements of monkeys. J Neurophysiol 67:625-638
[Abstract/Free Full Text] .- Gonshor A, Melvill-Jones G (1973) Changes of human vestibulo-ocular response induced by vision-reversal during head rotation. J Physiol (Lond) 234:102-103P.
- Highstein SM (1973) Synaptic linkage in the vestibulo-ocular and cerebello-vestibular pathways to the 6th nucleus in the rabbit. Exp Brain Res 17:301-314[Web of Science][Medline].
- Highstein SM, Partsalis A, Arikan R (1997) Role of the Y-group of the vestibular nuclei and flocculus of the cerebellum in motor learning of the vestibulo-ocular reflex. Prog Brain Res 114:383-397[Web of Science][Medline].
- Ito M (1972) Neural design of the cerebellar motor control system. Brain Res 40:81-84[Web of Science][Medline].
- Ito M (1982) Cerebellar control of the vestibulo-ocular reflex-around the flocculus hypothesis. Annu Rev Neurosci 5:275-298[Web of Science][Medline].
- Ito M (1993) Cerebellar flocculus hypothesis. Nature 363:24-25[Medline].
- Ito M, Shiida T, Yagi N, Yamamoto M (1974) The cerebellar modification of rabbit's horizontal vestibulo-ocular reflex induced by sustained head rotation combined with visual stimulation. Proc Jpn Acad 50:85-89.
- Ito M, Nisimaru N, Yamamoto M (1977) Specific patterns of neuronal connexions involved in the control of the rabbit's vestibulo-ocular reflexes by the cerebellar flocculus. J Physiol (Lond) 265:833-854
[Abstract/Free Full Text] .- Ito M, Sakurai M, Tongroach P (1982) Climbing fibre induced depression of mossy fibre responsiveness and glutamate sensitivity of cerebellar Purkinje cells. J Physiol (Lond) 324:113-124
[Abstract/Free Full Text] .- Judge SJ, Richmond BJ, Chu FC (1980) Implantation of magnetic search coils for measurement of eye position: an improved method. Vision Res 20:535-538[Web of Science][Medline].
- Kettner RE, Mahamud S, Leung H-C, Sitkoff N, Houk JC, Peterson BW, Barto AG (1997) Prediction of complex two-dimensional trajectories by a cerebellar model of smooth pursuit eye movement. J Neurophysiol 77:2115-2130
[Abstract/Free Full Text] .- Klopf AH (1989) Classical conditioning phenomena predicted by a drive-reinforcement model of neuronal function. In: Neural models of plasticity: experimental and theoretical approaches (Byrne JH, Berry WO, eds), pp 104-132. San Diego: Academic.
- Lev-Ram V, Makings LR, Keitz PF, Kao JPY, Tsien RY (1995) Long-term depression in cerebellar Purkinje neurons results from coincidence of nitric oxide and depolarization-induced Ca2+ transients. Neuron 15:407-415[Web of Science][Medline].
- Lev-Ram V, Jiang T, Wood J, Lawrence DS, Tsien RY (1997) Synergies and coincidence requirements between NO, cGMP, and Ca2+ in the induction of cerebellar long-term depression. Neuron 18:1025-1038[Web of Science][Medline].
- Lisberger SG (1994) Neural basis for motor learning in the vestibuloocular reflex of primates. III. Computational and behavioral analysis of the sites of learning. J Neurophysiol 72:974-998
[Abstract/Free Full Text] .- Lisberger SG (1998) Cerebellar LTD: a molecular mechanism of behavioral learning? Cell 92:701-704[Web of Science][Medline].
- Lisberger SG, Fuchs AF (1978a) Role of primate flocculus during rapid behavioral modification of vestibuloocular reflex. I. Purkinje cell activity during visually guided horizontal smooth-pursuit eye movements and passive head rotation. J Neurophysiol 41:733-763
[Abstract/Free Full Text] .- Lisberger SG, Fuchs AF (1978b) Role of primate flocculus during rapid behavioral modification of vestibuloocular reflex. II. Mossy fiber firing patterns during horizontal head rotation and eye movement. J Neurophysiol 41:764-777
[Abstract/Free Full Text] .- Lisberger SG, Pavelko TA (1986) Vestibular signals carried by pathways subserving plasticity of the vestibulo-ocular reflex in monkeys. J Neurosci 6:346-354[Abstract].
- Lisberger SG, Pavelko TA (1988) Brain stem neurons in modified pathways for motor learning in the primate vestibulo-ocular reflex. Science 242:771-773
[Abstract/Free Full Text] .- Lisberger SG, Sejnowski TJ (1992) Motor learning in a recurrent network model based on the vestibulo-ocular reflex. Nature 360:159-161[Medline].
- Lisberger SG, Westbrook LE (1985) Properties of visual inputs that initiate horizontal smooth pursuit eye movements in monkeys. J Neurosci 5:1662-1673[Abstract].
- Lisberger SG, Evinger C, Johanson GW, Fuchs AF (1981) Relationship between eye acceleration and retinal image velocity during foveal smooth pursuit in man and monkey. J Neurophysiol 46:229-249
[Free Full Text] .- Lisberger SG, Pavelko TA, Broussard DM (1994a) Responses during eye movements of brain stem neurons that receive monosynaptic inhibition from the flocculus and ventral paraflocculus in monkeys. J Neurophysiol 72:909-927
[Abstract/Free Full Text] .- Lisberger SG, Pavelko TA, Broussard DM (1994b) Neural basis for motor learning in the vestibuloocular reflex of primates. I. Changes in the responses of brain stem neurons. J Neurophysiol 72:928-953
[Abstract/Free Full Text] .- Lisberger SG, Pavelko TA, Bronte-Stewart HM, Stone LS (1994c) Neural basis for motor learning in the vestibuloocular reflex of primates. II. Changes in the responses of horizontal gaze velocity Purkinje cells in the cerebellar flocculus and ventral paraflocculus. J Neurophysiol 72:954-973
[Abstract/Free Full Text] .- Luebke AE, Robinson DA (1994) Gain changes of the cat's vestibulo-ocular reflex after flocculus deactivation. Exp Brain Res 98:379-390[Web of Science][Medline].
- Marr D (1969) A theory of cerebellar cortex. J Physiol (Lond) 202:437-470
[Abstract/Free Full Text] .- Mauk MD, Garcia KS, Medina JF, Steele PM (1998) Does cerebellar LTD mediate motor learning? Toward a resolution without a smoking gun. Neuron 20:359-362[Web of Science][Medline].
- Miles FA, Fuller JH (1974) Adaptive plasticity in the vestibulo-ocular responses of the rhesus monkey. Brain Res 80:512-516[Web of Science][Medline].
- Miles FA, Lisberger SG (1981) Plasticity in the vestibulo-ocular reflex: a new hypothesis. Annu Rev Neurosci 4:273-299[Web of Science][Medline].
- Miles FA, Fuller JH, Braitman DJ, Dow BM (1980a) Long-term adaptive changes in primate vestibuloocular reflex. III. Electrophysiological observations in flocculus of normal monkeys. J Neurophysiol 43:1437-1476
[Free Full Text] .- Miles FA, Braitman DJ, Dow BM (1980b) Long-term adaptive changes in primate vestibuloocular reflex. IV. Electrophysiological observations in flocculus of adapted monkeys. J Neurophysiol 43:1477-1493
[Free Full Text] .- Moore JW, Desmond JE, Berthier NE, Blazis DE, Sutton RS, Barto AG (1986) Simulation of the classically conditioned nicitating membrane response by a neuron-like adaptive element: response topography, neuronal firing, and interstimulus intervals. Behav Brain Res 21:143-154[Web of Science][Medline].
- Partsalis AM, Zhang Y, Highstein SM (1995) Dorsal Y group in the squirrel monkey. II. Contribution of the cerebellar flocculus to neuronal responses in normal and adapted animals. J Neurophysiol 73:632-649
[Abstract/Free Full Text] .- Pastor AM, de la Cruz RR, Baker R (1994) Cerebellar role in adaptation of goldfish vestibuloocular reflex. J Neurophysiol 72:1383-1394
[Abstract/Free Full Text] .- Raymond JL, Lisberger SG (1996) Behavioral analysis of signals that guide learned changes in the amplitude and dynamics of the vestibulo-ocular reflex. J Neurosci 16:7791-7802
[Abstract/Free Full Text] .- Raymond JL, Lisberger SG (1997) Multiple subclasses of Purkinje cells in the primate floccular complex provide similar signals to guide learning in the vestibulo-ocular reflex. Learn Mem 3:503-518
[Abstract/Free Full Text] .- Raymond JL, Lisberger SG, Mauk MD (1996) The cerebellum: a neuronal learning machine? Science 272:1126-1131[Abstract].
- Sakurai M (1987) Synaptic modification of parallel fibre-Purkinje cell transmission in in vitro guinea-pig cerebellar slices. J Physiol (London) 394:463-480
[Abstract/Free Full Text] .- Schreurs BG, Alkon DL (1993) Rabbit cerebellar slice analysis of long-term depression and its role in classical conditioning. Brain Res 631:235-240[Web of Science][Medline].
- Schreurs BG, Alkon DL (1996) Pairing-specific long-term depression of Purkinje cell excitatory postsynaptic potentials results from a classical conditioning procedure in the rabbit cerebellar slice. J Neurophysiol 75:1051-1060
[Abstract/Free Full Text] .- Shimazu H, Precht W (1966) Inhibition of central vestibular neurons from the contralateral labyrinth and its mediating pathway. J Neurophysiol 29:467-492
[Free Full Text] .- Stone LS, Lisberger SG (1990) Visual responses of Purkinje cells in the cerebellar flocculus during smooth-pursuit eye movements in monkeys. II. Complex spikes. J Neurophysiol 63:1262-1275
[Abstract/Free Full Text] .- Watanabe E (1984) Neuronal events correlated with long-term adaptation of the horizontal vestibulo-ocular reflex in the primate flocculus. Brain Res 297:169-174[Web of Science][Medline].
- Wurtz RH (1969) Visual receptive fields of striate cortex neurons in awake monkeys. Neurosci Res 32:727-742.
Copyright © 1998 Society for Neuroscience 0270-6474/98/18219112-18$05.00/0
This article has been cited by other articles:
L. R. Squire
Memory and Brain Systems: 1969-2009
J. Neurosci., October 14, 2009; 29(41): 12711 - 12716.
[Full Text] [PDF]
R. Ramachandran and S. G. Lisberger
Neural Substrate of Modified and Unmodified Pathways for Learning in Monkey Vestibuloocular Reflex
J Neurophysiol, October 1, 2008; 100(4): 1868 - 1878.
[Abstract] [Full Text] [PDF]
D. V. Sarkisov and S. S.-H. Wang
Order-Dependent Coincidence Detection in Cerebellar Purkinje Neurons at the Inositol Trisphosphate Receptor
J. Neurosci., January 2, 2008; 28(1): 133 - 142.
[Abstract] [Full Text] [PDF]
S. Ono and M. J. Mustari
Horizontal Smooth Pursuit Adaptation in Macaques After Muscimol Inactivation of the Dorsolateral Pontine Nucleus (DLPN)
J Neurophysiol, November 1, 2007; 98(5): 2918 - 2932.
[Abstract] [Full Text] [PDF]
V. S. Huang and R. Shadmehr
Evolution of Motor Memory During the Seconds After Observation of Motor Error
J Neurophysiol, June 1, 2007; 97(6): 3976 - 3985.
[Abstract] [Full Text] [PDF]
R. R. Kimpo and J. L. Raymond
Impaired Motor Learning in the Vestibulo-Ocular Reflex in Mice with Multiple Climbing Fiber Input to Cerebellar Purkinje Cells
J. Neurosci., May 23, 2007; 27(21): 5672 - 5682.
[Abstract] [Full Text] [PDF]
H. Straka, J. C. Beck, A. M. Pastor, and R. Baker
Morphology and Physiology of the Cerebellar Vestibulolateral Lobe Pathways Linked to Oculomotor Function in the Goldfish
J Neurophysiol, October 1, 2006; 96(4): 1963 - 1980.
[Abstract] [Full Text] [PDF]
J. C. Beck, P. Rothnie, H. Straka, S. L. Wearne, and R. Baker
Precerebellar Hindbrain Neurons Encoding Eye Velocity During Vestibular and Optokinetic Behavior in the Goldfish
J Neurophysiol, September 1, 2006; 96(3): 1370 - 1382.
[Abstract] [Full Text] [PDF]
R. Soetedjo and A. F. Fuchs
Complex spike activity of purkinje cells in the oculomotor vermis during behavioral adaptation of monkey saccades.
J. Neurosci., July 19, 2006; 26(29): 7741 - 7755.
[Abstract] [Full Text] [PDF]
R. R. Kimpo, E. S. Boyden, A. Katoh, M. C. Ke, and J. L. Raymond
Distinct Patterns of Stimulus Generalization of Increases and Decreases in VOR Gain
J Neurophysiol, November 1, 2005; 94(5): 3092 - 3100.
[Abstract] [Full Text] [PDF]
R. Ramachandran and S. G. Lisberger
Normal Performance and Expression of Learning in the Vestibulo-Ocular Reflex (VOR) at High Frequencies
J Neurophysiol, April 1, 2005; 93(4): 2028 - 2038.
[Abstract] [Full Text] [PDF]
M. R. Cohen, G. W. Meissner, R. J. Schafer, and J. L. Raymond
Reversal of Motor Learning in the Vestibulo-Ocular Reflex in the Absence of Visual Input
Learn. Mem., September 1, 2004; 11(5): 559 - 565.
[Abstract] [Full Text] [PDF]
S. Ono, V. E. Das, and M. J. Mustari
Gaze-Related Response Properties of DLPN and NRTP Neurons in the Rhesus Macaque
J Neurophysiol, June 1, 2004; 91(6): 2484 - 2500.
[Abstract] [Full Text] [PDF]
D. M. Broussard and C. D. Kassardjian
Learning in a Simple Motor System
Learn. Mem., March 1, 2004; 11(2): 127 - 136.
[Abstract] [Full Text] [PDF]
V. Lev-Ram, S. B. Mehta, D. Kleinfeld, and R. Y. Tsien
Reversing cerebellar long-term depression
PNAS, December 23, 2003; 100(26): 15989 - 15993.
[Abstract] [Full Text] [PDF]
L. Madelain and R. J. Krauzlis
Effects of Learning on Smooth Pursuit During Transient Disappearance of a Visual Target
J Neurophysiol, August 1, 2003; 90(2): 972 - 982.
[Abstract] [Full Text] [PDF]
C. Sekirnjak, B. Vissel, J. Bollinger, M. Faulstich, and S. du Lac
Purkinje Cell Synapses Target Physiologically Unique Brainstem Neurons
J. Neurosci., July 16, 2003; 23(15): 6392 - 6398.
[Abstract] [Full Text] [PDF]
J. E. Roy and K. E. Cullen
Brain Stem Pursuit Pathways: Dissociating Visual, Vestibular, and Proprioceptive Inputs During Combined Eye-Head Gaze Tracking
J Neurophysiol, July 1, 2003; 90(1): 271 - 290.
[Abstract] [Full Text] [PDF]
S. Ono, V. E. Das, and M. J. Mustari
Role of the Dorsolateral Pontine Nucleus in Short-Term Adaptation of the Horizontal Vestibuloocular Reflex
J Neurophysiol, May 1, 2003; 89(5): 2879 - 2885.
[Abstract] [Full Text] [PDF]
R. A. Clendaniel, D. M. Lasker, and L. B. Minor
Differential Adaptation of the Linear and Nonlinear Components of the Horizontal Vestibuloocular Reflex in Squirrel Monkeys
J Neurophysiol, December 1, 2002; 88(6): 3534 - 3540.
[Abstract] [Full Text] [PDF]
K. Yamamoto, Y. Kobayashi, A. Takemura, K. Kawano, and M. Kawato
Computational Studies on Acquisition and Adaptation of Ocular Following Responses Based on Cerebellar Synaptic Plasticity
J Neurophysiol, March 1, 2002; 87(3): 1554 - 1571.
[Abstract] [Full Text] [PDF]
Y. Hirata and S. M. Highstein
Acute Adaptation of the Vestibuloocular Reflex: Signal Processing by Floccular and Ventral Parafloccular Purkinje Cells
J Neurophysiol, May 1, 2001; 85(5): 2267 - 2288.
[Abstract] [Full Text] [PDF]
M. Kahlon and S. G. Lisberger
Changes in the Responses of Purkinje Cells in the Floccular Complex of Monkeys After Motor Learning in Smooth Pursuit Eye Movements
J Neurophysiol, December 1, 2000; 84(6): 2945 - 2960.
[Abstract] [Full Text] [PDF]
A. L. Babalian and P.-P. Vidal
Floccular Modulation of Vestibuloocular Pathways and Cerebellum-Related Plasticity: An In Vitro Whole Brain Study
J Neurophysiol, November 1, 2000; 84(5): 2514 - 2528.
[Abstract] [Full Text] [PDF]
T. Belton and R. A. McCrea
Role of the Cerebellar Flocculus Region in Cancellation of the VOR During Passive Whole Body Rotation
J Neurophysiol, September 1, 2000; 84(3): 1599 - 1613.
[Abstract] [Full Text] [PDF]
This Article Abstract 
Full Text (PDF) Submit an eLetter Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager 
Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (68) Citing Articles via Google Scholar Google Scholar Articles by Raymond, J. L. Articles by Lisberger, S. G. Search for Related Content PubMed PubMed Citation Articles by Raymond, J. L. Articles by Lisberger, S. G.
|
|
|
|
 |
|