Search

The visual and vestibular systems begin functioning early in life. However, it is unclear whether young infants perceive the dynamic world based on the retinal coordinate (egocentric reference frame) or the environmental coordinate (allocentric reference frame) when they encounter incongruence between frames of reference due to changes in body position. In this study, we performed the habituation-dishabituation procedure to assess novelty detection in a visual display, and a change in body position was included between the habituation and dishabituation phases in order to test whether infants dishabituate to the change in stimulus on the retinal or environmental coordinate. Twenty infants aged 3-4 months were placed in the right-side-down position (RSDp) and habituated to an animated human-like character that walked horizontally in the environmental frame of reference. Subsequently, their body position was changed in the roll plane. Ten infants were repositioned to the upright position (UPp) and the rest, to the RSDp after rotation. In the test phase, the displays that were spatially identical to those shown in the habituation phase and 90 degrees rotated displays were alternately presented, and visual preference was examined. The results revealed that infants looked longer at changes in the display on the retinal coordinate than at changes in the display on the environmental coordinate. This suggests that changes in body position from lying to upright produced incongruence of the egocentric and allocentric reference frames for perception of dynamic visual displays and that infants may rely more on the egocentric reference frame.

Saccular and utricular organs are essential for postural stability and gaze control. Although saccular and utricular inputs are known to terminate on vestibular neurons, few previous studies have precisely elucidated the origin of these inputs. We investigated the saccular and utricular inputs to single vestibular neurons in whole vestibular nuclei of decerebrated cats. Postsynaptic potentials were recorded from vestibular neurons after electrical stimulation of the saccular and utricular nerves. Ascending and descending axonal projections were examined by stimulating the oculomotor/trochlear nuclei and the cervical segment of the spinal cord, respectively. After each experiment, locations of recorded neurons were identified. The recorded neurons (140) were classified into vestibulo-spinal (79), vestibulo-oculo-spinal (9), and vestibulo-ocular (3) neurons based on antidromic responses; 49 other vestibular neurons were unidentified. The majority of recorded neurons were mainly located in the lateral vestibular nucleus. Most of the otolith-activated vestibular nuclei neurons seemed to participate in vestibulospinal reflexes. Of the total 140 neurons recorded, approximately one third (51) received saccular and utricular inputs (convergent neurons). The properties of these 51 convergent neurons were further investigated. Most (33/51) received excitatory postsynaptic potentials (EPSPs) after saccular and utricular nerve stimulation. These results implied that most of the convergent neurons in this study additively coded mixed information for vertical and horizontal linear acceleration. Based on the latencies of convergent neurons, we found that an early integration process for vertical and horizontal linear acceleration existed at the second-order level.

Connections from the otolithic organs to sternocleidomastoid (SCM) motoneurons were studied in 20 decerebrate cats. The electrical stimulation was selective for the saccular or the utricular nerves. Postsynaptic potentials were recorded from antidromically identified SCM motoneurons; these muscles participate mainly in neck rotation and flexion. Partial transections of the brainstem at the level of the obex were performed to identify the possible pathway from the otolithic organs to the SCM motoneurons. Saccular or utricular nerve stimulation mainly evoked inhibitory postsynaptic potentials (IPSPs) in the ipsilateral SCM motoneurons. Some of the sacculus-induced IPSPs were preceded by small-amplitude excitatory PSPs (EPSPs). The latencies of the PSPs ranged from 1.8 to 3.1 ms after saccular nerve stimulation and from 1.7 to 2.8 ms after utricular nerve stimulation, indicating that most of the ipsilateral connections were disynaptic. In the contralateral SCM motoneurons, saccular nerve stimulation had no or faint effects, whereas utricular nerve stimulation evoked EPSPs in about two-thirds of neurons, and no visible PSPs in about one-third of neurons. The latencies of the EPSPs ranged from 1.5 to 2.0 ms, indicating the disynaptic connection. Thus, the results suggest a difference between the two otolithic innervating patterns of SCM motoneurons. After transection of the medial vestibulospinal tract (MVST), saccular nerve stimulation did not evoke IPSPs at all in ipsilateral SCM motoneurons, but some (11/40) neurons showed small-amplitude EPSPs. Most (24/33) of the utricular-activated IPSPs disappeared after transection, whereas the other 9 neurons still indicated IPSPs. In the contralateral SCM motoneurons, no utricular-activated EPSPs were recorded after transection. These MVST transection results suggest that most of the otolith-SCM pathways are located in the MVST at the obex level. However, the results also suggest the possibility that other otolith-SCM pathways exist at the obex level.

In our daily lives, we can appropriately perform movements on the earth, suggesting that the central nervous system takes into account gravitational forces that act on our bodies during the movements. Recently, gravitational forces have been observed to generate the direction-dependent differences in the spatial properties of the kinematics of prehension movements. However, little is known about how gravitational forces affect the temporal properties of the kinematics of these movements. In this study, we tried to elucidate the gravitational effects on the temporal properties of the kinematics of movements by comparing upward (against gravity) and downward (with gravity) movements. As a result, we found the direction-dependent differences in temporal kinematics in both the reaching and grasping components of movements. For the reaching component, a shorter acceleration time was observed for the upward movements compared to the downward movements. For the grasping component, participants opened their hands earlier and faster for the upward movements than for the downward movements. These direction-dependent differences in the temporal kinematics suggested that the central nervous system takes into account and takes advantage of gravitational effects in the motor plans and controls of vertical prehension movements.

We studied the axonal projections of vestibulospinal neurons activated from the posterior semicircular canal. The axonal projection level, axonal pathway, and location of the vestibulospinal neurons originating from the PC were investigated in seven decerebrated cats. Selective electrical stimulation was applied to the PC nerve, and extracellular recordings in the vestibular nuclei were performed. The properties of the PC nerve-activated vestibulospinal neurons were then studied. To estimate the neural pathway in the spinal cord, floating electrodes were placed at the ipsilateral (i) and contralateral (c) lateral vestibulospinal tract (LVST) and medial vestibulospinal tract (MVST) at the C1/C2 junction. To elucidate the projection level, floating electrodes were placed at i-LVST and MVST at the C3, T1, and L3 segments in the spinal cord. Collision block test between orthodromic inputs from the PC nerve and antidromic inputs from the spinal cord verified the existence of the vestibulospinal neurons in the vestibular nuclei. Most (44/47) of the PC nerve-activated vestibulospinal neurons responded to orthodromic stimulation to the PC nerve with a short (

We examined whether otolith-activated second- and third-order vestibular nucleus neurons received commissural inhibition from the contralateral otolithic macula oriented in the same geometric plane. For this purpose we performed intracellular recording in vestibular nucleus neurons after stimulation of the ipsi- and contralateral utricular and saccular nerves. More than half (41/72) of the utricular-activated second-order vestibular nucleus neurons received commissural inhibition from the contralateral utricular nerve. The remaining neurons (31/72) showed no visible response to contralateral utricular nerve stimulation. About half (17/36) of utricular-activated third-order neurons also received commissural inhibition from the contralateral utricular nerve. Approximately 10% (7/67) of saccular-activated second-order vestibular neurons received polysynaptic commissural inhibition, whereas 16% (11/67) received commissural facilitation. The majority (49/67) of saccular second-order vestibular neurons, and almost all (22/23) third-order neurons, showed no visible response to stimulation of the contralateral saccular nerve. The present findings suggest that many utricular-activated vestibular nucleus neurons receive commissural inhibition, which may provide a mechanism for increasing the sensitivity of vestibular neurons to horizontal linear acceleration and lateral tilt of the head. Commissural inhibition in the saccular system was less prominent than in the utricular system.

Properties of otolith inputs to vestibulocerebellar neurons were investigated in 14 adult cats. In the vestibular nuclei, we recorded single-unit activities that responded orthodromically after stimulation of the utricular and/or saccular nerves and antidromically after stimulation of the cerebellum (uvula-nodulus and anterior vermis). Descending axonal projections to the spinal cord were also examined by antidromic stimulation of the caudal end of the C1 segment. Forty-seven otolith-activated neurons that projected to the uvula-nodulus were recorded. Thirteen (28%) of the 47 neurons received convergent inputs from the utriculus and sacculus. The remaining 34 (72%) vestibular neurons were non-convergent neurons: 18 (38%) received utricular input alone, and 16 (34%) received saccular input alone. Most (35/47) vestibulocerebellar neurons were located in the descending vestibular nucleus and only one of these projected to the spinal cord. Seven of the 47 vestibulocerebellar neurons were located in the lateral vestibular nucleus and most of these neurons projected to the spinal cord. The remaining neurons were located in group X (two neurons) and the superior vestibular nucleus (three neurons). In a different series of experiments, 37 otolith-activated vestibular neurons were tested to determine whether they projected to the uvula-nodulus and/or the anterior vermis. Nineteen of the 37 neurons projected to the anterior vermis, 13/37 projected to the uvula-nodulus, and 5/37 projected to both. The utricular and/or saccular nerve-activated vestibulocerebellar neurons projected to not only the uvula-nodulus, but also to the anterior vermis. In summary, the results of this study showed that vestibular neurons receiving inputs from the utriculus and/or sacculus projected to the cerebellar cortex. This indirect otolith-cerebellar pathway terminated both in the anterior lobe and in the uvula/nodulus.

The convergence of the posterior semicircular canal (PC) and utricular (UT) inputs in single vestibular nuclei neurons was studied intracellularly in decerebrate cats. A total of 160 vestibular neurons were orthodromically activated by selective stimulation of the PC and the UT nerve and classified according to whether or not they were antidromically activated from the spinal cord and oculomotor nuclei into vestibulospinal (VS), vestibulooculospinal (VOS), vestibuloocular (VO), and unidentified vestibular neurons. Fifty-three (33%) of 160 vestibular neurons received convergent inputs from both the PC and UT nerves. Seventy-nine (49%) vestibular neurons responded to PC inputs alone, and 28 (18%) neurons received inputs only from the UT nerve. Of 53 convergent neurons, 8 (15%) were monosynaptically excited from both nerves. Thirty-five (66%) received monosynaptic excitatory inputs from the PC nerve and polysynaptic excitatory or inhibitory inputs from the UT nerve, or vice versa. Approximately one-third of VS and VOS neurons received convergent inputs. A majority of the VS neurons descended to the spinal cord through the lateral vestibulospinal tract, while almost all the VOS neurons descended to the spinal cord through the medial vestibulospinal tract. The convergent neurons were found in all vestibular nuclei but more in the lateral nucleus and descending nucleus. The VS neurons were more numerous than VO neurons or VOS neurons.

Convergence between posterior canal (PC) and saccular (SAC) inputs in single vestibular nuclei neurons was investigated in decerebrated cats. Postsynaptic potentials were recorded intracellularly after selective stimulation of the SAC and PC nerves. Stimulation of either the SAC or PC nerve orthodromically activated 143 vestibular nuclei neurons. Of these, 61 (43%) were antidromically activated by stimulation of the C1–C2 junction, 14 (10%) were antidromically activated by stimulation of the oculomotor or trochlear nucleus, and 14 (10%) were antidromically activated by stimulation of both the oculomotor or trochlear nucleus and the spinal cord. Fifty-four (38%) neurons were not activated by stimulation of either or both. We named these neurons vestibulospinal (VS), vestibulo-ocular (VO), vestibulo-oculo-spinal (VOS) and vestibular (V) neurons, respectively. Both PC and SAC inputs converged in 47 vestibular nuclei neurons (26 VS, 2 VO, 6 VOS and 13 V neurons). Of these, 19 received monosynaptic excitatory inputs from both nerves. This input pattern was frequently seen in VS neurons. Approximately half of the convergent VS neurons descended to the spinal cord through the lateral vestibulospinal tract. The remaining half and all the convergent VOS neurons descended to the spinal cord through the medial vestibulospinal tract. Most of the convergent neurons were located in the lateral nucleus or descending nucleus.

The otolith system contributes to the vestibulo-ocular reflexes (VOR) when the head moves linearly in the horizontal plane or tilts relative to gravity. The saccules are thought to detect predominantly accelerations along the gravity vector. Otolith-induced vertical eye movements following vertical linear accelerations are attributed to the saccules. However, information on the neural circuits of the sacculo-ocular system is limited, and the effects of saccular inputs on extraocular motoneurons remain unclear. In the present study, synaptic responses to saccular-nerve stimulation were recorded intracellularly from identified motoneurons of all twelve extraocular muscles. Experiments were successfully performed in eleven cats. Individual motoneurons of the twelve extraocular muscles--the bilateral superior recti (SR), inferior recti (IR), superior obliques (SO), inferior obliques (IO), lateral recti (LR), and medial recti (MR) were identified antidromically following bipolar stimulation of their respective nerves. The saccular nerve was selectively stimulated by a pair of tungsten electrodes after removing the utricular nerve and the ampullary nerves of the semicircular canals. Stimulus intensities were determined from the stimulus-response curves of vestibular N1 field potentials in order to avoid current spread. Intracellular recordings were performed from 129 extraocular motoneurons. The majority of the neurons showed no response to saccular-nerve stimulation. In 17 (30%) of 56 extraocular motoneurons related to vertical eye movements (bilateral SR and IR), depolarizing and/or hyperpolarizing postsynaptic potentials (PSPs) were observed in response to saccular-nerve stimulation. The latencies of PSPs ranged from 2.3 to 8.9 ms, indicating that the extraocular motoneurons received neither monosynaptic nor disynaptic inputs from saccular afferents. The majority of the latencies of the depolarization, including depolarization-hyperpolarization, were in the range of 2.3-3.3 ms. Latencies of hyperpolarizations were typically longer than those of depolarizations. Only one contralateral SO motoneuron of 43 recorded oblique extraocular motoneurons (bilateral SO and IO) showed a depolarization-hyperpolarization in response to saccular-nerve stimulation at a latency of 2.5 ms. None of 30 recorded horizontal extraocular motoneurons (bilateral LR and MR) responded to stimulation of the saccular nerve. The neural linkage in the sacculo-ocular system is relatively weak in comparison to the utriculo-ocular and sacculo-collic systems, suggesting that the role of the sacculo-ocular system in stabilizing eye position may be reduced when compared with utriculo-ocular and semi-circular canal-ocular reflexes.