Optimal response planes and canal convergence in secondary neurons in vestibular nuclei of alert cats

Adaptation of spatio-temporal convergent properties in central vestibular neurons in monkeys

The spatio-temporal convergent (STC) response occurs in central vestibular cells when dynamic and static inputs are activated. The functional significance of STC behavior is not fully understood. Whether STC is a property of some specific central vestibular neurons, or whether it is a response that can be induced in any neuron at some frequencies is unknown. It is also unknown how the change in orientation of otolith polarization vector (orientation adaptation) affects STC behavior. A new complex model, that includes inputs with regular and irregular discharges from both canal and otolith afferents, was applied to experimental data to determine how many convergent inputs are sufficient to explain the STC behavior as a function of frequency and orientation adaptation. The canal-otolith and otolith-only neurons were recorded in the vestibular nuclei of three monkeys. About 42% (11/26 canal-otolith and 3/7 otolith-only) neurons showed typical STC responses at least at one frequency before orientation adaptation. After orientation adaptation in side-down head position for 2 h, some canal-otolith and otolith-only neurons altered their STC responses. Thus, STC is a property of weights of the regular and irregular vestibular afferent inputs to central vestibular neurons which appear and/or disappear based on stimulus frequency and orientation adaptation. This indicates that STC properties are more common for central vestibular neurons than previously assumed. While gravity-dependent adaptation is also critically dependent on stimulus frequency and orientation adaptation, we propose that STC behavior is also linked to the neural network responsible for localized contextual learning during gravity-dependent adaptation.

Vasovagal oscillations and vasovagal responses produced by the vestibulo-sympathetic reflex in the rat

Sinusoidal galvanic vestibular stimulation (sGVS) induces oscillations in blood pressure (BP) and heart rate (HR), i.e., vasovagal oscillations, as well as transient decreases in BP and HR, i.e., vasovagal responses, in isoflurane-anesthetized rats. We determined the characteristics of the vasovagal oscillations, assessed their role in the generation of vasovagal responses, and determined whether they could be induced by monaural as well as by binaural sGVS and by oscillation in pitch. Wavelet analyses were used to determine the power distributions of the waveforms. Monaural and binaural sGVS and pitch generated vasovagal oscillations at the frequency and at twice the frequency of stimulation. Vasovagal oscillations and vasovagal responses were maximally induced at low stimulus frequencies (0.025-0.05 Hz). The oscillations were attenuated and the responses were rarely induced at higher stimulus frequencies. Vasovagal oscillations could occur without induction of vasovagal responses, but vasovagal responses were always associated with a vasovagal oscillation. We posit that the vasovagal oscillations originate in a low frequency band that, when appropriately activated by strong sympathetic stimulation, can generate vasovagal oscillations as a precursor for vasovagal responses and syncope. We further suggest that the activity responsible for the vasovagal oscillations arises in low frequency, otolith neurons with orientation vectors close to the vertical axis of the head. These neurons are likely to provide critical input to the vestibulo-sympathetic reflex to increase BP and HR upon changes in head position relative to gravity, and to contribute to the production of vasovagal oscillations and vasovagal responses and syncope when the baroreflex is inactivated.

Spatial orientation of the angular vestibulo-ocular reflex (aVOR) after semicircular canal plugging and canal nerve section

We investigated spatial responses of the aVOR to small and large accelerations in six canal-plugged and lateral canal nerve-sectioned monkeys. The aim was to determine whether there was spatial adaptation after partial and complete loss of all inputs in a canal plane. Impulses of torques generated head thrusts of ≈ 3,000°/s². Smaller accelerations of ≈ 300°/s² initiated the steps of velocity (60°/s). Animals were rotated about a spatial vertical axis while upright (0°) or statically tilted fore-aft up to ± 90°. Temporal aVOR yaw and roll gains were computed at every head orientation and were fit with a sinusoid to obtain the spatial gains and phases. Spatial gains peaked at ≈ 0° for yaw and ≈ 90° for roll in normal animals. After bilateral lateral canal nerve section, the spatial yaw and roll gains peaked when animals were tilted back ≈ 50°, to bring the intact vertical canals in the plane of rotation. Yaw and roll gains were identical in the lateral canal nerve-sectioned monkeys tested with both low- and high-acceleration stimuli. The responses were close to normal for high-acceleration thrusts in canal-plugged animals, but were significantly reduced when these animals were given step stimuli. Thus, high accelerations adequately activated the plugged canals, whereas yaw and roll spatial aVOR gains were produced only by the intact vertical canals after total loss of lateral canal input. We conclude that there is no spatial adaptation of the aVOR even after complete loss of specific semicircular canal input.

Coordinate transformations and sensory integration in the detection of spatial orientation and self-motion: from models to experiments

An accurate internal representation of our current motion and orientation in space is critical to navigate in the world and execute appropriate action. The force of gravity provides an allocentric frame of reference that defines one's motion relative to inertial (i.e., world-centered) space. However, movement in this environment also introduces particular motion detection problems as our internal linear accelerometers, the otolith organs, respond identically to either translational motion or changes in head orientation relative to gravity. According to physical principles, there exists an ideal solution to the problem of distinguishing between the two as long as the brain also has access to accurate internal estimates of angular velocity. Here, we illustrate how a nonlinear integrative neural network that receives sensory signals from the vestibular organs could be used to implement the required computations for inertial motion detection. The model predicts several distinct cell populations that are comparable with experimentally identified cell types and accounts for a number of previously unexplained characteristics of their responses. A key model prediction is the existence of cell populations that transform head-referenced rotational signals from the semicircular canals into spatially referenced estimates of head reorientation relative to gravity. This chapter provides an overview of how addressing the problem of inertial motion estimation from a computational standpoint has contributed to identifying the actual neuronal populations responsible for solving the tilt-translation ambiguity and has facilitated the interpretation of neural response properties.

Adaptation of orientation vectors of otolith-related central vestibular neurons to gravity

Behavioral experiments indicate that central pathways that process otolith-ocular and perceptual information have adaptive capabilities. Because polarization vectors of otolith afferents are directly related to the electro-mechanical properties of the hair cell bundle, it is unlikely that they change their direction of excitation. This indicates that the adaptation must take place in central pathways. Here we demonstrate for the first time that otolith polarization vectors of canal-otolith convergent neurons in the vestibular nuclei have adaptive capability. A total of 10 vestibular-only and vestibular-plus-saccade neurons were recorded extracellularly in two monkeys before and after they were in side-down positions for 2 h. The spatial characteristics of the otolith input were determined from the response vector orientation (RVO), which is the projection of the otolith polarization vector, onto the head horizontal plane. The RVOs had no specific orientation before animals were in side-down positions but moved toward the gravitational axis after the animals were tilted for extended periods. Vector reorientations varied from 0 to 109 degrees and were linearly related to the original deviation of the RVOs from gravity in the position of adaptation. Such reorientation of central polarization vectors could provide the basis for changes in perception and eye movements related to prolonged head tilts relative to gravity or in microgravity.

Organization of dye-coupled cerebellar granule cells labeled from afferent vestibular and dorsal root fibers in the frogRana esculenta

Application of neurobiotin to the nerves of individual labyrinthine organs and dorsal root fibers of limb-innervating segments of the frog resulted in labeling of granule cells in the cerebellum showing a significant overlap with a partial segregation in the related areas of termination. In different parts of the cerebellum, various combinations of different canal and otolith organ-related granule cells have been discerned. The difference in the extension of territories of vertical canals vs. horizontal canals may reflect their different involvement in the vestibuloocular and vestibulospinal reflex. Dye-coupled cells related to the lagenar and saccular neurons were localized in more rostral parts of the cerebellum, whereas cells of the utricle were represented only in its caudal half. This separation is supportive of the dual function of the lagena and the saccule. The territories of granule cells related to the cervical and lumbar segments of the spinal cord were almost completely separated along the rostrocaudal axis of cerebellum, whereas their territories were almost entirely overlapping in the mediolateral and ventrodorsal directions. The partial overlap of labyrinthine organ-related and dorsal root fiber-related granule cells are suggestive of a convergence of sensory modalities involved in the sense of balance. We propose that the afferent input of vestibular and proprioceptive fibers mediated by gap junctions to the cerebellar granule cells subserve one of the possible morphological correlates of a very rapid modification of the motor activity in the vestibulocerebellospinal neuronal circuit.

Spatial properties of central vestibular neurons

We studied the spatial characteristics of 45 vestibular-only (VO) and 12 vestibular-plus-saccade (VPS) neurons in two cynomolgus monkeys using angular rotation and static tilt. The purpose was to determine the contribution of canal and otolith-related inputs to central vestibular neurons whose activity is associated with the central velocity storage integrator. Lateral canal-related neurons responded maximally during vertical axis rotation when the head was tilted 25 +/- 6 and 22 +/- 3 degrees forward relative to the axis of rotation in the two animals, and vertical canal-related neurons responded maximally with the head tilted back 63+/- 5 and 57 +/- 7 degrees . The origin of the vertical canal-related input was verified by rotation about a spatial horizontal axis. Thirty-one percent of cells received input in a single canal plane. Sixty-seven percent of canal-related cells received otolith input, 31% of vertical canal neurons had lateral canal input, and 43% of lateral canal neurons had vertical canal input. Twenty percent of neurons had convergent input from the lateral canals, the vertical canals, and the otolith organs. Some VO and VPS cells had spatial-temporal convergent (STC) properties; more of these cells had STC properties at lower frequencies of rotation. Thus VO and VPS neurons associated with velocity storage receive a broad range of convergent inputs from each portion of the vestibular labyrinth. This convergence could provide the basis for gravity-dependent eye velocity orientation induced through velocity storage.

Spatial properties of central vestibular neurons of monkeys after bilateral lateral canal nerve section

Thirty-seven neurons were recorded in the superior vestibular nucleus (SVN) of two cynomolgus monkeys 1-2 yr after bilateral lateral canal nerve section to test whether the central neurons had spatially adapted for the loss of lateral canal input. The absence of lateral canal function was verified with eye movement recordings. The relation of unit activity to the vertical canals was determined by oscillating the animals about a horizontal axis with the head in various orientations relative to the axis of rotation. Animals were also oscillated about a vertical axis while upright or tilted in pitch. In the second test, the vertical canals are maximally activated when the animals are tilted back about -50 degrees from the spatial upright and the lateral canals when the animals are tilted forward about 30 degrees . We reasoned that if central compensation occurred, the head orientation at which the response of the vertical canal-related neurons was maximal should be shifted toward the plane of the lateral canals. No lateral canal-related units were found after nerve section, and vertical canal-related units were found only in SVN not in the rostral medial vestibular nucleus. SVN canal-related units were maximally activated when the head was tilted back at -47 +/- 17 and -50 +/- 12 degrees (means +/- SD) in the two animals, close to the predicted orientation of the vertical canals. This indicated that spatial adaptation of vertical canal-related vestibular neurons had not occurred. There were substantial neck and/or otolith-related inputs activating the vertical canal-related neurons in the nerve-sectioned animals, which could have contributed to oculomotor compensation after nerve section.

Electromyographic activity of dorsal neck muscles in squirrel monkeys during rotations in an upright or upside down posture

Electromyographic (EMG) activity was recorded from occipitoscapularis, semispinalis, and splenius neck muscles in five alert squirrel monkeys during 0.25-Hz rotations about horizontal axes oriented at 22.5 degrees intervals, including pitch, roll, and intermediate axes. The animals were oriented in either upright or upside down posture. In the upright posture, all monkeys exhibited compensatory EMG activity with maximal activation during rotations about axes between pitch in the pitch forward direction and contralaterally directed roll. Response timing varied across animals with EMG peaks ranging from near pitch forward head velocity to near pitch forward head position. When the head was upside down, response dynamics and directionality were altered to varying degrees in different monkeys. The greatest change in response to head inversion was seen in the monkey that had response phases closest to head position, the least in the animal with phases closest to head velocity. The monkey with EMG response peaks closest to position phase showed nearly 180 degrees inversion of responses when the head was upside down, suggesting that in this monkey a righting reflex mediated by utricular signals was activated in the upside down posture. The monkey with EMG response peaks closest to velocity phase may have lacked a righting response and exhibited only a canal-mediated compensatory vestibulocervical reflex in both upright and upside down postures. The results suggest that reflex contraction of neck muscles in response to passive head rotation includes an interplay of compensatory and righting responses that varies from animal to animal.

Properties of cerebellar fastigial neurons during translation, rotation, and eye movements

The most medial of the deep cerebellar nuclei, the fastigial nucleus (FN), receives sensory vestibular information and direct inhibition from the cerebellar vermis. We investigated the signal processing in the primate FN by recording single-unit activities during translational motion, rotational motion, and eye movements. Firing rate modulation during horizontal plane translation in the absence of eye movements was observed in all non-eye-movement-sensitive cells and 26% of the pursuit eye-movement-sensitive neurons in the caudal FN. Many non-eye-movement-sensitive cells recorded in the rostral FN of three fascicularis monkeys exhibited convergence of signals from both the otolith organs and the semicircular canals. At low frequencies of translation, the majority of these rostral FN cells changed their firing rates in phase with head velocity rather than linear acceleration. As frequency increased, FN vestibular neurons exhibited a wide range of response dynamics with most cells being characterized by increasing phase leads as a function of frequency. Unlike cells in the vestibular nuclei, none of the rostral FN cells responded to rotational motion alone, without simultaneously exhibiting sensitivity to translational motion. Modulation during earth-horizontal axis rotation was observed in more than half (77%) of the neurons, although with smaller gains than during translation. In contrast, only 47% of the cells changed their firing rates during earth-vertical axis rotations in the absence of a dynamic linear acceleration stimulus. These response properties suggest that the rostral FN represents a main processing center of otolith-driven information for inertial motion detection and spatial orientation.

An Integrative Neural Network for Detecting Inertial Motion and Head Orientation

The ability to navigate in the world and execute appropriate behavioral responses depends critically on the contribution of the vestibular system to the detection of motion and spatial orientation. A complicating factor is that otolith afferents equivalently encode inertial and gravitational accelerations. Recent studies have demonstrated that the brain can resolve this sensory ambiguity by combining signals from both the otoliths and semicircular canal sensors, although it remains unknown how the brain integrates these sensory contributions to perform the nonlinear vector computations required to accurately detect head movement in space. Here, we illustrate how a physiologically relevant, nonlinear integrative neural network could be used to perform the required computations for inertial motion detection along the interaural head axis. The proposed model not only can simulate recent behavioral observations, including a translational vestibuloocular reflex driven by the semicircular canals, but also accounts for several previously unexplained characteristics of central neural responses such as complex otolith–canal convergence patterns and the prevalence of dynamically processed otolith signals. A key model prediction, implied by the required computations for tilt–translation discrimination, is a coordinate transformation of canal signals from a head-fixed to a spatial reference frame. As a result, cell responses may reflect canal signal contributions that cannot be easily detected or distinguished from otolith signals. New experimental protocols are proposed to characterize these cells and identify their contributions to spatial motion estimation. The proposed theoretical framework makes an essential first link between the computations for inertial acceleration detection derived from the physical laws of motion and the neural response properties predicted in a physiologically realistic network implementation.

The response of vestibulo-ocular reflex pathways to electrical stimulation after canal plugging

The vestibulo-ocular reflex (VOR) allows clear vision during head movements by generating compensatory eye movements. Its response to horizontal rotation is reduced after one horizontal semicircular canal is plugged, but recovers partially over time. The majority of VOR interneurons contribute to the shortest VOR pathway, the so-called three-neuron arc, which includes only two synapses in the brainstem. After a semicircular canal is plugged, transmission of signals by the three-neuron arc originating from the undamaged side may be altered during recovery. We measured the oculomotor response to single current pulses delivered to the vestibular labyrinth of alert cats between 9 h and 1 month after plugging the contralateral horizontal canal. The same response was also measured after motor learning induced by continuously-worn telescopes (optically induced motor learning). Optically induced learning did not change the peak velocity of the evoked eye movement (PEEV) significantly but, after a canal plug, the PEEV increased significantly, reaching a maximum during the first few post-plug days and then decreasing. VOR gain also showed transient changes during recovery. Because the PEEV occurred early in the eye movement evoked by a current pulse, we think the observed increase in PEEV represented changes in transmission by the three-neuron arc. Sham surgery did not result in significant changes in the response to electrical stimulation or in VOR gain. Our data suggest that different pathways and processes may underlie optically induced motor learning and recovery from plugging of the semicircular canals.

Proprioception from a spinocerebellar perspective

This review explores how proprioceptive sensory information is organized at spinal cord levels as it relates to a sense of body position and movement. The topic is considered in an historical context and develops a different framework that may be more in tune with current views of sensorimotor processing in other central nervous system structures. The dorsal spinocerebellar tract (DSCT) system is considered in detail as a model system that may be considered as an end point for the processing of proprioceptive sensory information in the spinal cord. An analysis of this system examines sensory processing at the lowest levels of synaptic connectivity with central neurons in the nervous system. The analysis leads to a framework for proprioception that involves a highly flexible network organization based in some way on whole limb kinematics. The functional organization underlying this framework originates with the biomechanical linkages in the limb that establish functional relationships among the limb segments. Afferent information from limb receptors is processed further through a distributed neural network in the spinal cord. The result is a global representation of hindlimb parameters rather than a muscle-by-muscle or joint-by-joint representation.

Spatial distribution of semicircular canal nerve evoked monosynaptic response components in frog vestibular nuclei

Most second-order vestibular neurons receive a canal-specific monosynaptic excitation, although the central projections of semicircular canal afferents overlap extensively. This remarkable canal specificity prompted us to study the spatial organization of evoked field potentials following selective stimulation of individual canal nerves. Electrically evoked responses in the vestibular nuclei were mapped systematically in vitro. Constructed activation maps were superimposed on a cytoarchitectonically defined anatomical map. The spatial activation maps for pre- and postsynaptic response components evoked by stimulation of a given canal nerve were similar. Activation maps for monosynaptic inputs from different canals tended to show a differential distribution of their peak amplitudes, although the overlap was considerable. Anterior vertical canal signals peaked in the superior vestibular nucleus, posterior vertical canal signals peaked in the descending and in the dorsal part of the lateral vestibular nucleus, whereas horizontal canal signals peaked in the descending and in the ventral part of the lateral vestibular nucleus. A similar, differential but overlapping, spatial organization of the canal inputs was described also for other vertebrates, suggesting a crude but rather conservative topographical organization of semicircular canal nerve projections within the vestibular nuclei. Differences in the precision of topological representations between vestibular and other sensory modalities are discussed.

Vertical Purkinje cells of the monkey floccular lobe: simple-spike activity during pursuit and passive whole body rotation

To understand how the simian floccular lobe is involved in vertical smooth pursuit eye movements and the vertical vestibuloocular reflex (VOR), we examined simple-spike activity of 70 Purkinje (P) cells during pursuit eye movements and passive whole body rotation. Fifty-eight cells responded during vertical and 12 during horizontal pursuit. We classified P cells as vertical gaze velocity (VG) if their modulation occurred for movements of both the eye (during vertical pursuit) and head (during pitch VOR suppression) with the modulation during one less than twice that of the other and was less during the target-fixed-in-space condition (pitch VOR X1) than during pitch VOR suppression. VG P cells constituted only a minority of vertical P cells (19%). Other vertical P cells that responded during pitch VOR suppression were classified as vertical eye and head velocity (V/) P cells (48%), regardless of the synergy of their response direction during smooth pursuit and VOR suppression. Vertical P cells that did not respond during pitch VOR suppression but did respond during rotation in vertical planes other than pitch were classified as off-pitch V/ P cells (33%). The mean eye-velocity and eye-position sensitivities of the three types of vertical P cells were similar. One-third (2/7 VG, 2/11 V/, 6/13 off-pitch V/), in addition, showed eye position sensitivity during saccade-free fixations. Maximal vestibular activation directions (MADs) were examined during VOR suppression by applying vertical whole body rotation with the monkeys oriented in different vertical planes. The MADs for VG P cells and V/ P cells with eye and vestibular sensitivity in the same direction were distributed near the pitch plane, suggesting convergence of bilateral anterior canal inputs. In contrast, MADs of off-pitch V/ P cells and V/ P cells with oppositely directed eye and vestibular sensitivity were shifted toward the roll plane, suggesting convergence of anterior and posterior canal inputs of the same side. Unlike horizontal G P cells, the modulation of many VG and V/ P cells when the target was fixed in space (pitch VOR X1) was not well predicted by the linear addition of their modulations during vertical pursuit and pitch VOR suppression. These results indicate that the populations of vertical and horizontal eye-movement P cells in the floccular lobe have markedly different discharge properties and therefore may be involved in different kinds of processing of vestibular-oculomotor interactions.

Fastigial nucleus activity during different frequencies and orientations of vertical vestibular stimulation in the monkey

Neurons in the rostral part of the fastigial nucleus (FN) respond to vestibular stimulation but are not related to eye movements. To understand the precise role of these vestibular-only neurons in the central processing of vestibular signals, unit activity in the FN of alert monkeys (Macaca mulatta) was recorded. To induce vestibular stimulation, the monkey was rotated sinusoidally around an earth-fixed horizontal axis at stimulus frequencies between 0.06 (+/-15 degrees) and 1.4 Hz (+/-7.5 degrees). During stimulation head orientation was changed continuously, allowing for roll, pitch, and intermediate planes of orientation. At a frequency of 0.6 Hz, 59% of the neurons had an optimal response orientation (ORO) and a null response (i.e., no modulation) 90 degrees apart. The phase of neuronal response was constant except for a steep shift of 180 degrees around the null response. This group I response is compatible with a semicircular canal input, canal convergence, or a single otolith input. Several other features indicated more complex responses, including spatiotemporal convergence (STC). 1) For 35% of the responses at 0.6 Hz, phase changes were gradual with different orientations. Fifteen percent of these had a null response (group II), and 20% showed only a minimal response but no null response (group III). The remaining responses (6%), classified as group IV, were characterized by a constant sensitivity at different orientations in most instances. 2) For the vast majority of neurons, the stimulus frequency determined the response group, i.e., an individual neuron could show a group I response at one frequency and a group II (III or IV) response at another frequency. 3) ORO changed with frequency by >45 degrees for 44% of the neurons. 4) Although phase changes at different frequencies were close to head velocity (+/-45 degrees ) or head position (+/-45 degrees ) for most neurons, they exceeded 90 degrees for 29% of the neurons between 0.1 and 1.0 Hz. In most cases, this was a phase advance. The change in sensitivity with change in frequency showed a similar pattern for all neurons; the average sensitivity increased from 1.24 imp. s-1. deg-1 at 0.1 Hz to 2.97 imp. s-1. deg-1 at 1.0 Hz. These data demonstrate that only an analysis based on measurements at different frequencies and orientations reveals a number of complex features. They moreover suggest that for the vast majority of neurons several sources of canal and otolith information interact at this central stage of vestibular information processing.

Otolith processing in the deep cerebellar nuclei

To investigate the otolith contribution to the responses of "vestibular only" neurons in the rostral fastigial nucleus (FN), single-unit activity was recorded in the alert monkey with the head fixed during static and dynamic stimulation (+/- 15 deg, 0.06-1.4 Hz) around an earth-fixed horizontal axis. Head orientation could be altered allowing for roll, pitch, and intermediate planes of orientation. For the vast majority of neurons a response vector orientation (RVO) with an optimal response and a null-response at a head orientation 90 deg apart could be determined. Presumably more than 30% of the vestibular only neurons had an otolith input, as indicated by responses to static tilt, head-position-related activity, large phase changes (> 100 deg) of neuronal activity between 0.06 and 1.4 Hz, changes of the RVO at different frequencies and complex responses (spatio-temporal convergence). Thus, neurons in FN reflecting an otolith or a combined canal-otolith input are much more common than up to now thought. Vestibular-only neurons are most likely involved in vestibulospinal mechanisms. Their precise functional role has yet to be determined.

A theory of geometric constraints on neural activity for natural three-dimensional movement

Although the orientation of an arm in space or the static view of an object may be represented by a population of neurons in complex ways, how these variables change with movement often follows simple linear rules, reflecting the underlying geometric constraints in the physical world. A theoretical analysis is presented for how such constraints affect the average firing rates of sensory and motor neurons during natural movements with low degrees of freedom, such as a limb movement and rigid object motion. When applied to nonrigid reaching arm movements, the linear theory accounts for cosine directional tuning with linear speed modulation, predicts a curl-free spatial distribution of preferred directions, and also explains why the instantaneous motion of the hand can be recovered from the neural population activity. For three-dimensional motion of a rigid object, the theory predicts that, to a first approximation, the response of a sensory neuron should have a preferred translational direction and a preferred rotation axis in space, both with cosine tuning functions modulated multiplicatively by speed and angular speed, respectively. Some known tuning properties of motion-sensitive neurons follow as special cases. Acceleration tuning and nonlinear speed modulation are considered in an extension of the linear theory. This general approach provides a principled method to derive mechanism-insensitive neuronal properties by exploiting the inherently low dimensionality of natural movements.

Discharge characteristics of vestibular saccade neurons in alert monkeys

We previously described a class of neurons, located in and around the interstitial nucleus of Cajal of the cat, that discharged during vestibular stimulation and before saccades. We called these neurons vestibular saccade neurons (VSNs). In the present study, we characterized similar neurons in the monkey. These neurons discharged before vertical saccades and during vertical vestibular stimulation as well as vertical smooth pursuit. Like cat VSNs, the discharge metrics of these VSNs were poorly related to saccade metrics and showed only occasional, weak sensitivity to eye position. They discharged most intensely (on-direction) for movements that were either upward or downward, and their on-directions were consistent during pitch and pursuit but not for eye position. For saccades, the correlation coefficient of number of spikes and vertical saccade size varied from 0.08 to 0.90 with a mean of approximately 0.6. The average sensitivity (i.e., slope) of the number of spikes and vertical saccade size linear regression was 0.3 +/- 0.2 spike/deg. Average correlations between peak discharge rate and peak saccade velocity and between burst duration and saccade duration were 0.5 and 0.4; sensitivities were 0.2 +/- 0.2 spike per s per deg/s and 0.6 +/- 0.5 ms/ms, respectively. Average vestibular sensitivities during 0.5 Hz, +/-10 degrees sinusoidal pitch while the animals suppressed their vestibular ocular reflex were 0.97 spike/s per deg/s for up VSNs and 0.66 spike/s per deg/s for down VSNs. The average static position sensitivity for the population of 39 VSNs tested was 0.55 spike/s per deg. The average gain for VSNs tested during 0.5 Hz, +/-10 degrees sinusoidal smooth pursuit tracking was 1.4 spike/s per deg/s. As we could not identify analogous neurons in the region of the monkey ponto-medullary junction, we conclude that horizontal on-direction VSNs do not exist in the monkey. We discuss a possible functional role for VSNs and similar neurons described in previous studies and conclude that these neurons are most likely involved with the process of neural integration (in a mathematical sense) of velocity-coded inputs from a variety of oculomotor subsystems and are not a pivotal element in saccade generation.

Processing of spatial information by floccular and non-floccular target neurons in the alert cat

In five alert chronically-prepared cats we studied the response to sinusoidal 3-D whole body rotation of well-isolated vestibular nucleus neurons which were tested for monosynaptic input from the vestibular labyrinth, direct projection to the oculomotor nucleus and, in some cases, inhibition from cerebellar flocculus stimulation. Neurons directly inhibited by flocculus stimulation had significantly greater horizontal-vertical semicircular canal signal convergence than did neurons not inhibited by flocculus stimulation.

Interdependence of spatial properties and projection patterns of medial vestibulospinal tract neurons in the cat

Activity of vestibular nucleus neurons with axons in the ipsi- or contralateral medial vestibulospinal tract was studied in decerebrate cats during sinusoidal, whole-body rotations in many planes in three-dimensional space. Antidromic activation of axon collaterals distinguished between neurons projecting only to neck segments from those with collaterals to C6 and/or oculomotor nucleus. Secondary neurons were identified by monosynaptic activation after labyrinth stimulation. A three-dimensional maximum activation direction vector (MAD) summarized the spatial properties of 151 of 169 neurons. The majority of secondary neurons (71%) terminated above the C6 segment. Of these, 43% had ascending collaterals to the oculomotor nucleus (VOC neurons), and 57% did not (VC neurons). The majority of VOC and VC neurons projected contralaterally and ipsilaterally, respectively. Most C6-projecting neurons could not be activated from oculomotor nucleus (V-C6 neurons) and projected primarily ipsilaterally. All VO-C6 neurons projected contralaterally. The distributions of MADs for secondary neurons with different projection patterns were different. Most VOC (84%) and contralaterally projecting VC (91%) neurons had MADs close to the activation vector of a semicircular canal pair, compared with 54% of ipsilaterally projecting VC (i-VC) and 39% of V-C6 neurons. Many i-VC (44%) and V-C6 (48%) neurons had responses suggesting convergent input from horizontal and vertical canal pairs. Horizontal and vertical gains were comparable for some, making it difficult to assign a primary canal input. MADs consistent with vertical-vertical canal pair convergence were less common. Type II yaw or type II roll responses were seen for 22% of the i-VC neurons, 68% of the V-C6 neurons, and no VOC cells. VO-C6 neurons had spatial properties between those of VOC and V-C6 neurons. These results suggest that secondary VOC neurons convey semicircular canal pair signals to both ocular and neck motor centers, perhaps linking eye and head movements. Secondary VC and V-C6 neurons carry more processed signals, possibly to drive neck and forelimb reflexes more selectively. Two groups of secondary i-VC neurons exhibited vertical-horizontal canal convergence similar to that present on neck muscles. The vertical-vertical canal convergence present on many neck muscles, however, was not present on medial vestibulospinal neurons. Spatial transformations achieved by the vestibulocollic reflex may occur in part on secondary neurons but further combination of canal signals must take place to generate compensatory muscle activity.

Excitatory and inhibitory inputs from saccular afferents to single vestibular neurons in the cat

Connections from saccular afferents to vestibular neurons were studied by means of intracellular recordings of excitatory (E) and inhibitory (I) postsynaptic potentials (PSPs) in vestibular neurons after focal stimulation of the saccular macula in decerebrated cats. Focal stimulation was given to the saccular macula in two ways, in which the polarity of stimulus current via a pair of electrodes was changed. In group A, one of the electrodes was inserted into the ventral and the other into the dorsal edge of the saccular macula. The focal stimulation was across the striola so that the reversal of morphological polarization in hair cells was bridged by the pulse stimulus. In 22/36 vestibular neurons tested, the stimulation of the saccular macula evoked monosynaptic (</=1.2 ms) EPSPs, including EPSP-IPSP sequences, with one polarity of stimulation, and disynaptic (>/=1.5 ms) IPSPs when the polarity of the stimulus current was changed. In 14/36 neurons, the response pattern was the same regardless of the stimulus polarity; EPSPs (12/36) or IPSPs (2/36). In group B, a pair of electrodes was inserted into the dorsal edge of the saccular macula, so that the striola was not bridged by the current stimulus. In all of the vestibular neurons tested, the response pattern was always the same regardless of the polarity: mono- (22/31) and disynaptic (3/31) EPSPs or disynaptic IPSPs (6/31). In addition, the saccular nerve was stimulated after removing the macula in some cats (group C). The stimulation of the saccular nerve evoked EPSPs in 62 vestibular neurons (including EPSP-IPSP sequences in 31 neurons) and IPSPs in 19 vestibular neurons. Convergence between the saccular nerve and other vestibular nerves was studied by the intracellular recording of PSPs. Fifty-six percent (18/32) of the saccular-activated neurons had excitatory and/or inhibitory potentials evoked after stimulation of the utricular nerve and the horizontal and anterior semicircular canal nerves, and 44% (19/43) of the neurons received inputs from the posterior semicircular canal nerve. The results support the hypothesis that saccular afferents from one population of hair cells activate vestibular neurons monosynaptically and that afferents from another population of hair cells located on the opposite side of the striola appear to project to the same vestibular neurons disynaptically via inhibitory interneurons. Neural circuits from saccular afferents to vestibular neurons, which we term cross-striolar inhibition, thus may provide a mechanism for increasing the sensitivity to vertical linear acceleration. The circuit described is provided not only with high sensitivity but also with input noise-resistant characteristics.

Canal-specific excitation and inhibition of frog second-order vestibular neurons

Second-order vestibular neurons (secondary VNs) were identified in the in vitro frog brain by their monosynaptic excitation following electrical stimulation of the ipsilateral VIIIth nerve. Ipsilateral disynaptic inhibitory postsynaptic potentials were revealed by bath application of the glycine antagonist strychnine or of the gamma-aminobutyric acid-A (GABA(A)) antagonist bicuculline. Ipsilateral disynaptic excitatory postsynaptic potentials (EPSPs) were analyzed as well. The functional organization of convergent monosynaptic and disynaptic excitatory and inhibitory inputs onto secondary VNs was studied by separate electrical stimulation of individual semicircular canal nerves on the ipsilateral side. Most secondary VNs (88%) received a monosynaptic EPSP exclusively from one of the three semicircular canal nerves; fewer secondary VNs (10%) were monosynaptically excited from two semicircular canal nerves; and even fewer secondary VNs (2%) were monosynaptically excited from each of the three semicircular canal nerves. Disynaptic EPSPs were present in the majority of secondary VNs (68%) and originated from the same (homonymous) semicircular canal nerve that activated a monosynaptic EPSP in a given neuron (22%), from one or both of the other two (heteronymous) canal nerves (18%), or from all three canal nerves (28%). Homonymous activation of disynaptic EPSPs prevailed (74%) among those secondary VNs that exhibited disynaptic EPSPs. Disynaptic inhibitory postsynaptic potentials (IPSPs) were mediated in 90% of the tested secondary VNs by glycine, in 76% by GABA, and in 62% by GABA as well as by glycine. These IPSPs were activated almost exclusively from the same semicircular canal nerve that evoked the monosynaptic EPSP in a given secondary VN. Our results demonstrate a canal-specific, modular organization of vestibular nerve afferent fiber inputs onto secondary VNs that consists of a monosynaptic excitation from one semicircular canal nerve followed by disynaptic excitatory and inhibitory inputs originating from the homonymous canal nerve. Excitatory and inhibitory second-order (secondary) vestibular interneurons are envisaged to form side loops that mediate spatially similar but dynamically different signals to secondary vestibular projection neurons. These feedforward side loops are suited to adjust the dynamic response properties of secondary vestibular projection neurons by facilitating or disfacilitating phasic and tonic input components.

Rostral fastigial nucleus activity in the alert monkey during three-dimensional passive head movements

The fastigial nucleus (FN) receives vestibular information predominantly from Purkinje cells of the vermis. FN in the monkey can be divided in a rostral part, related to spinal mechanisms, and a caudal part with oculomotor functions. To understand the role of FN during movements in space, single-unit activity in alert monkeys was recorded during passive three-dimensional head movements from rostral FN. Seated monkeys were rotated sinusoidally around a horizontal earth-fixed axis (vertical stimulation) at different orientations 15 degrees apart (including roll, pitch, vertical canal plane and intermediate planes). In addition, sinusoidal rotations around an earth-vertical axis (yaw stimulus) included different roll and pitch positions (+/-10 degrees, +/-20 degrees). The latter positions were also used for static stimulation. One hundred fifty-eight neurons in two monkeys were modulated during the sinusoidal vertical search stimulation. The vast majority showed a uniform response pattern: a maximum at a specific head orientation (response vector orientation) and a null response 90 degrees apart. Detailed analysis was obtained from 111 neurons. On the basis of their phase relation during dynamic stimulation and their response to static tilt, these neurons were classified as vertical semicircular canal related (n = 79, 71.2%) or otolith related (n = 25; 22.5%). Only seven neurons did not follow the usual response pattern and were classified as complex neurons. For the vertical canal-related neurons (n = 79) all eight major response vector orientations (ipsilateral or contralateral anterior canal, posterior canal, roll, and nose-down and nose-up pitch) were found in Fn on one side. Neurons with ipsilateral orientations were more numerous and on average more sensitive than those with contralateral orientations. Twenty-eight percent of the vertical canal-related neurons also responded to horizontal canal stimulation. None of the vertical canal-related neurons responded to static tilt. Otolith-related neurons (n = 25) had a phase relation close to head position and were considerably less numerous than canal-related neurons. Except for pitch, all other response vector orientations were found. Seventy percent of these neurons responding during dynamic stimulation also responded during static tilt. The sensitivity during dynamic stimulation was always higher than during static stimulation. Sixty-one percent of the otolith-related neurons responded also to horizontal canal stimulation. These results show that in FN, robust vestibular signals are abundant. Canal-related responses are much more common than otolith-related responses. Although for many canal neurons the responses can be related to single canal planes, convergence between vertical canals but also with horizontal canals is common.

Cross axis vestibulo-ocular reflex induced by pursuit training in alert monkeys

We examined the effect of conflicting vestibular and smooth pursuit information on the vestibulo-ocular reflex (VOR) in alert monkeys. Sinusoidal whole body rotation was applied either in the pitch or yaw plane while presenting a target spot that moved orthogonally to the rotation plane. The monkeys were rewarded for tracking the spot and eye movements induced by rotation alone were examined every 15-30 min in complete darkness. Orthogonal eye movement responses to rotation were not observed before training but appeared after about 30 min of training. The gains (eye/chair) of the orthogonal component increased up to 0.2 after 1-2 h and were largest at the training frequency and approximately in phase with the stimulus; phase advanced at lower frequencies and lagged at higher frequencies. Amplitude tuning was also demonstrated when examined using different amplitudes at a constant frequency after training. Two hours of training to fixate an earth-stationary spot projected onto a patterned visual background that moved orthogonally to the rotation plane during rotation, did not induce a cross axis response. These results indicate that pursuit training during VOR is effective in inducing cross-axis eye movement responses that are tuned to the metrics of the training stimulus.

Simple-spike activity of floccular Purkinje cells responding to sinusoidal vertical rotation and optokinetic stimuli in alert cats

To understand how the cerebellar flocculus is involved in the processing of semicircular canal signals in the vertical vestibulo-ocular reflex (VOR), we analyzed the simple-spike activity of floccular Purkinje (P) cells that was modulated by sinusoidal pitch rotation, and then analyzed their activity during presentation of sinusoidal vertical optokinetic stimuli in alert, head-fixed cats. The great majority of P cells also responded to optokinetic stimuli with peak discharge near peak stimulus velocity. Eighty percent of P cells that responded to both pitch and optokinetic stimuli showed increased activity when the directions of the resultant eye movements were the same. During rapid modification of the VOR induced by visual pattern movement, modulation amplitudes of the cells tested increased together with the eye velocity increase. Maximal activation directions of these cells studied during vertical rotation in many planes were near the vertical canal planes, similar to those in our previous studies. The remaining 20% of P cells showed increased discharge for the same direction of stimulus movement. These results suggest that the activity of the majority of pitch-responding P cells contains, at least partly, a vertical eye velocity component during presentation of vestibular or optokinetic stimuli in addition to canal inputs during pitch rotation.

Further evidence for the specific involvement of the flocculus in the vertical vestibulo-ocular reflex (VOR)

We examined the simple-spike activity of floccular Purkinje (P) cells during sinusoidal pitch rotation and vertical optokinetic stimuli in alert, head-fixed cats. The great majority of pitch-responding P cells also responded to optokinetic stimuli with increased activity when the directions of the resultant eye movements were the same. During rapid modification of the VOR induced by visual pattern movement, modulation amplitudes of the cells tested increased together with the eye velocity increase. Maximal activation directions of these cells studied during vertical rotation in many planes were near the vertical canal planes. These results suggest that the activity of the majority of pitch-responding P cells contains a vertical eye velocity component during vestibular or optokinetic stimuli in addition to canal inputs during pitch rotation.

Projections of the individual vestibular end-organs in the brain stem of the squirrel monkey

The central nervous system (CNS) projections of primary afferent neurons from individual vestibular receptors were studied using horseradish peroxidase (HRP) or biocytin labeling in 14 ears from 7 adult squirrel monkeys using the technique developed in the chinchilla (Lee et al., 1989, 1992). The specificity of labeling was verified by examining the location of the labeled fibers and cell bodies in the vestibular nerve and Scarpa's ganglion. Labeled fibers and cells were restricted to nerves and areas belonging to groups of cells in either the superior or the inferior ganglion of the vestibular nerve. In the vestibular nerve root, labeled primary afferent fibers also exhibited a receptor-dependent segregation at the entrance to the medulla. Fibers from the HSC and the SSC were found rostrally and those from the PSC and the SAC were found in the caudal area. The UTR fibers were situated intermediate between these two groups of fibers. (A bundle of fibers, probably vestibular efferents, was identified immediately rostrally and ventromedially to the UTR fibers.) The primary afferent fibers bifurcated into secondary ascending and descending fibers at the lateral border of the vestibular nuclei, forming a longitudinal rostrocaudal vestibular tract. The secondary fibers from individual end-organs occupied specific locations in the tract: the UTR fibers were dorsal to the SSC and the HSC fibers, PSC fibers were found most medially, and the SAC fibers occupied the lateralmost area. The secondary UTR fibers overlapped considerably with those of the SSC and the HSC. The orderly receptor-dependent segregation of fibers was more prominent in the descending tracts than in the ascending tracts. In the vestibular nuclei complex the location of the tertiary branches of various end-organs exhibited considerable overlap within the major vestibular nuclei (SN, superior nucleus; LN, lateral nucleus; MN, medial nucleus; DN, descending nucleus). There were still differences, however, in the projection pattern. Fibers from the SAC ran primarily in the lateral area, fibers from the SSC and the UTR were found ventromedially to the SAC fibers, and the HSC projected slightly medially to the fibers from the SSC. The PSC fibers projected most medially. The UTR and SAC sent numerous fibers to the cerebellum. Fibers from the semicircular canals projected through the rostrodorsal region of the SN and presumably also projected to the cerebellum. The precise termination of fibers was evaluated by studying the location of labeled boutons, which were identified in all major vestibular nuclei. Labeled boutons from all the receptors were in the rostral and central areas of the SN, and in the MN mainly in the rostral two-thirds. In the LN, boutons from all the receptors were in the rostroventral part, most of which were from the UTR and SAC. No labeled boutons were in the caudodorsal part of this nucleus. Labeled boutons in the DN primarily surrounded the descending tract fibers and were particularly prominent medially. In specimens in which superior vestibular nerve receptor organs were scratched vestibular efferent fibers were also labeled. These fibers traveled in the most ventral part of the vestibular nerve root and projected in the ventral aspect of the LN to labeled soma in the ipsilateral and contralateral brain stem. Specificity the in projection patterns of efferent fibers from different end-organs could not be ascertained.

Temporal response properties of lumbar-projecting vestibulospinal neurons to roll tilt in decerebrate cats

In decerebrate cats, rotation about the longitudinal axis of the animal, leading to sinusoidal stimulation of labyrinth receptors, produces a tonic contraction of limb extensors during side-down tilt (alpha responses) and of dorsal neck extensors during side-up tilt (beta responses). These changes in posture are mediated, at least in part, by lateral vestibular nucleus (LVN) neurons, with response characteristics to stimulation of macular and/or canal receptors that have so far been evaluated at the level of either unidentified vestibulospinal (VS) neurons or vestibulo-collic neurons projecting to the upper cervical cord. In the present study we investigated the dynamics of the responses of VS neurons projecting to the lumbosacral segments of the spinal cord to increasing frequencies of tilt (from 0.026 to 0.32 Hz, +/- 10 degrees). All the recorded units showed an average phase lead with respect to position of +25.6 +/- 5.5 degrees (SE) at the tilt frequency of 0.026 Hz. Most of these neurons (n = 32) were particularly activated during side-down tilt (alpha responses) and showed either a stable phase or an increase in phase lead of the responses with increasing frequency of tilt. At the tilt frequency of 0.026 Hz, the smaller the phase lead of the responses, the larger was the response gain. Moreover, the smaller the phase lead of the responses at that frequency of tilt, the smaller the increase in gain but the larger was the increase in lead of the responses obtained by increasing the stimulation frequency up to 0.32 Hz. Through this set of finely organized changes in unit response characteristics, the overall output of this population of neurons increased, while the phase angle of the responses reached the mean value of +64.9 +/- 2.6 degrees (SE), thus becoming more related to the velocity than to the positional signal. The remaining units (n = 7), which were mainly activated during side-up tilt (beta responses), displayed an increase in phase lag of the responses to increasing frequency of stimulation, which reached the mean value of -118.9 +/- 14.5 degrees (SE) at 0.32 Hz. The differences in the dynamic properties of these VS neurons projecting to the lumbosacral cord, with respect to those of previously recorded populations of VS neurons, are discussed.

Neural basis for eye velocity generation in the vestibular nuclei of alert monkeys during off-vertical axis rotation

Activity of "vestibular only" (VO) and "vestibular plus saccade" (VPS) units was recorded in the rostral part of the medial vestibular nucleus and caudal part of the superior vestibular nucleus of alert rhesus monkeys. By estimating the "null axes" of recorded units (n = 79), the optimal plane of activation was approximately the mean plane of reciprocal semicircular canals, i.e., lateral canals, left anterior-right posterior (LARP) canals or right anterior-left posterior (RALP) canals. All units were excited by rotation in a direction that excited a corresponding ipsilateral semicircular canal. Thus, they all displayed a "type I" response. With the animal upright, there were rapid changes in firing rates of both VO and VPS units in response to steps of angular velocity about a vertical axis. The units were bidirectionally activated during vestibular nystagmus (VN), horizontal optokinetic nystagmus (OKN), optokinetic after-nystagmus (OKAN) and off-vertical axis rotation (OVAR). The rising and falling time constants of the responses to rotation indicated that they were closely linked to velocity storage. There were differences between VPS and VO neurons in that activity of VO units followed the expected time course in response to a stimulus even during periods of drowsiness, when eye velocity was reduced. Firing rates of VPS units, on the other hand, were significantly reduced in the drowsy state. Lateral canal-related units had average firing rates that were linearly related to the bias or steady state level of horizontal eye velocity during OVAR over a range of +/- 60 deg/s. These units could be further divided into two classes according to whether they were modulated during OVAR. Non-modulated units (n = 5) were VO types and all modulated units (n = 5) were VPS types. There was no significant difference between the bias level sensitivities relative to eye velocity of the units with and without modulation (P > 0.05). The modulated units had no sustained change in firing rate in response to static head tilts and their phases relative to head position varied from unit to unit. The phase did not appear to be linked to the modulation of horizontal eye velocity during OVAR. The sensitivities of unit activity to eye velocity were similar during all stimulus modalities despite the different gains of eye velocity vs stimulus velocity during VN, OKN and OVAR. Therefore, VO and VPS units are likely to carry an eye velocity signal related to velocity storage.(ABSTRACT TRUNCATED AT 400 WORDS)

Response properties of gerbil otolith afferents to small angle pitch and roll tilts

The responses from isolated single otolith afferent fibers were obtained to small angle sinusoidal pitch and roll tilts in anesthetized gerbils. The stimulus directions that produced the maximum (response vector) and minimum response sensitivities were determined for each otolith afferent, with response vectors for the units being spread throughout the horizontal plane, similar to those reported for other species. A breadth of tuning measure was derived, with narrowly tuned neurons responding maximally to stimulation in one direction and minimally along an orthogonal ('null') direction. Most (approximately 80%) otolith afferents are narrowly tuned, however, some fibers were broadly tuned responding significantly to stimulations in any direction in the horizontal plane. The number of broadly tuned otolith afferents (approximately 20%) differs significantly from the more substantial number of broadly tuned vestibular nuclei neurons (88%) recently reported in rats.

Ultrastructural evidence that early synapse formation on central vestibular sensory neurons is independent of peripheral vestibular influences

Migration and early differentiation of neurons of the tangential vestibular nucleus of the chick take place between embryonic days 5 and 8. In the absence of primary vestibular afferents (otocyst-ablation), a previous light microscope study documented that early developmental events still occurred, but the neurons failed to complete differentiation and to survive. In order to understand why these neurons undergo normal early development, we have repeated the vestibular deafferentation paradigm followed by ultrastructural observations on these neurons. We found that the ultrastructural events associated with migration and differentiation in the deafferented tangential nucleus were essentially normal from 5 to 8 days. Most important, longitudinal fibers, presumably of central, nonvestibular origins, formed the first synapses at the same time and sequence as observed in normal embryos. Thus vestibular sensory neurons receive their first input from central fibers, initiating events in the formation of a central vestibular circuitry without the influence of peripheral vestibular fibers or endorgan.

The interstitial nucleus of Cajal in the midbrain reticular formation and vertical eye movement

Bilateral lesions of the midbrain reticular formation within, and in the close vicinity of, the interstitial nucleus of Cajal (INC) result in the severe impairment of the ability to hold eccentric vertical eye position after saccades, phase advance and decreased gain of the vestibulo-ocular reflex (VOR) induced by sinusoidal vertical rotation. In addition, the INC region of alert animals contains many burst-tonic and tonic neurons whose activity is closely correlated with vertical eye movement, not only during spontaneous saccades, but also during the VOR, smooth pursuit and optokinetic eye movements. Although their activity is closely related to these conjugate vertical eye movements, it is different from the oculomotor motor neuron activity. These results indicate that the INC region is involved in, and indispensable for, some aspects of eye position generation during vertical eye movement. Further comparison of INC neuron discharge with eye movements during two special conditions indicates that the INC region alone cannot produce eye position signals. First INC neuron discharge shows no response or an 80 degrees phase advance (close to the expected value if there is no integration) in the dark compared to the light during sinusoidal vertical linear acceleration in alert cats. Second, during rapid-eye-movement (REM) sleep, the discharge of INC neurons is no longer correlated with eye position. These results imply that the INC is not the entire velocity-to-position integrator, but that it has to work with other region(s) to perform the integration. A close functional linkage has been described between vertical-eye-movement-related neurons in the INC region and vestibulo-ocular relay neurons related to the vertical semicircular canals in the vestibular nuclei. It has been suggested that both are the major constituents of the common neural integrator circuits for vertical eye movements.

Excitation of the extraocular muscles in decerebrate cats during the vestibulo-ocular reflex in three-dimensional space

(1) Vestibulo-ocular reflex excitation of the six extraocular muscles was studied by recording their electromyographic activity in decerebrate cats during oscillations about horizontal and vertical axes, at frequencies from 0.07 to 4 Hz. Animals were oriented in many different positions and rotated about axes that lay in the horizontal, frontal, or sagittal planes defined by our coordinate system. (2) The strengths of modulation (gains) of the responses of all extraocular muscles were a sinusoidal function of the orientation of the rotation axis within a coordinate plane, and this function was nearly independent of rotation frequency. (3) The responses were used to determine an axis of maximal excitation for each of the extraocular muscles by the vestibulo-ocular reflex. Antagonistic muscle pairs were found to have best axes in nearly opposite directions, confirming their operation as pairs. (4) Excitation of the medial and lateral rectus could be explained by input from the paired horizontal semicircular canals, with essentially no convergent input from vertical canals. (5) Excitation of the vertical rectus and oblique muscles could be explained by convergent inputs from the vertical canals with little or no horizontal canal input.

Responses of frog vestibular neurons to combined temperature microstimulation of the semicircular canals

The temperature microstimulation of the semicircular canal (heat bursts lasting about 2 sec with a peak amplitude of 0.5-5.0 degrees C) can be regarded as an analog of angular acceleration acting within the cavity of the given canal; the combination of temperature microstimulations targeted simultaneously on several canals can serve as a physical model of accelerated rotation having complex spatial characteristics. The recording of the activity of neurons of the vestibular nuclei of the frog (n = 278) in response to temperature microstimulation of the canals showed that 80% of the neurons have inputs from one to two canals and only 20% from three to six canals. The distribution of laterality was characterized by the predominance of ipsilateral inputs: 201 neurons (72.3%) had only ipsilateral inputs; 14 neurons (5%) had only contralateral inputs; 63 neurons (22.7%) had both inputs. The ipsilateral horizontal (67.6%) and the posterior (61.9%) were the most effective inputs; while among the contralateral, the posterior (21.2%) canals were the most effective; the least effective were the contralateral horizontal (8.6%) and the anterior (5.0%) canals. The presence of latent canal inputs (both excitatory and inhibitory) which showed up in combination with effective inputs was demonstrated.

Spatial properties of second-order vestibulo-ocular relay neurons in the alert cat

Second-order vestibular nucleus neurons which were antidromically activated from the region of the oculomotor nucleus (second-order vestibuloocular relay neurons) were studied in alert cats during whole-body rotations in many horizontal and vertical planes. Sinusoidal rotation elicited sinusoidal modulation of firing rates except during rotation in a clearly defined null plane. Response gain (spike/s/deg/s) varied as a cosine function of the orientation of the cat with respect to a horizontal rotation axis, and phases were near that of head velocity, suggesting linear summation of canal inputs. A maximum activation direction (MAD) was calculated for each cell to represent the axis of rotation in three-dimensional space for which the cell responded maximally. Second-order vestibuloocular neurons divided into 3 non-overlapping populations of MADs, indicating primary canal input from either anterior, posterior or horizontal semicircular canal (AC, PC, HC cells). 80/84 neurons received primary canal input from ipsilateral vertical canals. Of these, at least 6 received input from more than one vertical canal, suggested by MAD azimuths which were sufficiently misaligned with their primary canal. In addition, 21/80 received convergent input from a horizontal canal, with about equal number of type I and type II yaw responses. 4/84 neurons were HC cells; all of them received convergent input from at least one vertical canal. Activity of many vertical second-order vestibuloocular neurons was also related to vertical and/or horizontal eye position. All AC and PC cells that had vertical eye position sensitivity had upward and downward on-directions, respectively. A number of PC cells had MADs centered around the MAD of the superior oblique muscle, and 2/3 AC cells recorded in the superior vestibular nucleus had MADs near that of the inferior oblique. Thus, signals with spatial properties appropriate to activate oblique eye muscles are present at the second-order vestibular neuron level. In contrast, none of the second-order vestibuloocular neurons had MADs near those of the superior or inferior rectus muscles. Signals appropriate to activate these eye muscles might be produced by combining signals from ipsilateral and contralateral AC neurons (for superior rectus) or PC neurons (for inferior rectus). Alternatively, less direct pathways such as those involving third or higher order vestibular or interstitial nucleus of Cajal neurons might play a crucial role in the spatial transformations between semicircular canals and vertical rectus eye muscles.

Distributed Parallel Processing in the Vestibulo-Oculomotor System

The mechanisms of eye-movement control are among the best understood in motor neurophysiology. Detailed anatomical and physiological data have paved the way for theoretical models that have unified existing knowledge and suggested further experiments. These models have generally taken the form of black-box diagrams (for example, Robinson 1981) representing the flow of hypothetical signals between idealized signal-processing blocks. They approximate overall oculomotor behavior but indicate little about how real eye-movement signals would be carried and processed by real neural networks. Neurons that combine and transmit oculomotor signals, such as those in the vestibular nucleus (VN), actually do so in a diverse, seemingly random way that would be impossible to predict from a block diagram. The purpose of this study is to use a neural-network learning scheme (Rumelhart et al. 1986) to construct parallel, distributed models of the vestibulo-oculomotor system that simulate the diversity of responses recorded experimentally from VN neurons.

Afferents and efferents of the vestibular nuclei: the necessity of context-specific interpretation

A synopsis of physiological and anatomical results is presented that leads to the conclusion that experimental data have to be interpreted in a context meaningful for the system investigated. For example, since there is an obvious spatial relationship between semicircular canals and extraocular muscles, the interdependence between the three-neurone-arc circuitry, and vestibular and visual signals follows quite naturally from a common geometry inherent in the sensory and motor periphery. It is emphasized that signals related to compensatory eye movements have to be interpreted within a vestibular/eye muscle frame of reference. By the same argument, when dealing with the head-neck movement system, the appropriate reference frame will have to be applied to arrive at a meaningful interpretation of related sensorimotor functions. Thus, in general terms, each system has to be interpreted within its own meaningful biological context.

Motor output to lateral rectus in cats during the vestibulo-ocular reflex in three-dimensional space

The motor output to the lateral rectus eye muscle was studied in decerebrate cats with electromyographic recordings and in alert cats with multi-unit and single neuron recordings from abducens nucleus. The axis of rotation that produced maximal excitation of the lateral rectus was calculated from responses to rotations in many different stimulus orientations, and was found to lie near the axis of the horizontal semicircular canals, but pitched slightly nose down from the canal axis (4.6 degrees). The results from decerebrate and alert cats were in agreement. The dynamics of lateral rectus activation were quite similar in all planes. Responses at high frequencies were in phase with rotation velocity and responses lagged toward position phase as frequency and velocity were decreased. Differences in decerebrate cat low frequency responses to rotations with and without a sinusoidal gravitational stimulus implicated an otolith input to lateral rectus.

Axonal trajectories of posterior canal-activated secondary vestibular neurons and their coactivation of extraocular and neck flexor motoneurons in the cat

Unit activities of 148 secondary vestibular neurons related to the posterior semicircular canal were recorded extracellularly in anesthetized cats. Axonal projections of these neurons were examined by their antidromic responses to stimulation of the excitatory target motoneurons of the contralateral (c-) inferior rectus muscle (IR) and bilateral (bi-) motoneuron pools of longus capitis muscles, neck flexors, in the C1 segment (C1LC). The neurons were classified into 4 groups according to their axonal projections. The first group of neurons, termed vestibulo-oculo-collic (VOC) neurons, sent axon collaterals both to the c-IR motoneuron pool and to the c-C1LC motoneuron pool. The majority of them (72%) were located in the descending nucleus. The second group of neurons were termed vestibuloocular (VO) neurons and sent their axons to the c-IR motoneuron pool but not to the cervical cord. Most of them (86%) were located in the medial nucleus. The third group of neurons, termed vestibulo-collic (contralateral) (VCc) neurons, sent axons to the c-C1LC motoneuron pool via the contralateral ventral funiculus but not to the oculomotor nuclei. They were mostly (75%) found in the descending nucleus. The last group of neurons were vestibulo-collic (ipsilateral) (VCi) neurons, which gave off axons to the ipsilateral (i-) C1LC motoneuron pool via the ipsilateral ventral funiculus but not to the oculomotor nuclei. One of them also sent an axon collateral to the c-C1LC motoneuron pool. The majority of them (74%) were located in the ventral part of the lateral nucleus. It was also observed in some of the VOC and VCi neurons that they produced unitary EPSPs in the c-C1LC and i-C1LC motoneurons, respectively. Their synaptic sites were estimated to be on the cell somata and/or proximal dendrites of the motoneurons.

The commissural inhibition on secondary vestibulo-ocular neurons in the vertical semicircular canal systems in the cat

The commissural inhibition on secondary vestibulo-ocular neurons (VOns) from the contralateral (c-) vertical canal system in the same geometric plane was studied in the anesthetized cat. The secondary VOns were identified by their orthodromic responses to stimulation of the ampullary nerves of the anterior (ACN) or posterior (PCN) semicircular canals and also by their antidromic responses to stimulation of the IIIrd and IVth nuclei. The majority of ACN-activated excitatory VOns in the descending and medial nuclei (32/36, 89%) and in the superior nucleus (20/23, 87%), received commissural inhibition from the c-PCN, while only few ACN-activated inhibitory VOns (3/35, 9%) in the superior nucleus received commissural inhibition from the c-PCN. On the other hand, all of the PCN-activated excitatory (50/50) and inhibitory (30/30) VOns in the vestibular nuclei received commissural inhibition following c-ACN stimulation.

Dynamics of adaptive change in vestibulo-ocular reflex direction. II. Sagittal plane rotations

Alert cats were rotated sinusoidally (0.25 Hz) in their sagittal plane while viewing optokinetic motion in their horizontal plane. Vertical and horizontal vestibulo-ocular reflex (VOR) was measured in the dark before and after 2 h of these adaptation stimuli in upright or onside orientation of the cat. Onside exposure produced maximal adaptive horizontal VOR at the training frequency. Upright exposure produced highest gain at lower frequencies. A cat with inoperative vertical canals adapted only to upright exposure. We conclude that in the presence of horizontal image rotation either vertical canal or otolith stimulation can produce adaptation in VOR direction and stimulation of both produces complex adaptation dynamics.