The vestibular primary afferent projection to the cerebellum of the rabbit was studied with retrograde and orthograde tracers. We injected individual lobules of the cerebellum with horseradish peroxidase (HRP) or wheat germ agglutinin-HRP (WGA-HRP). Following these injections the numbers of labeled and unlabeled cells in Scarpa's ganglion were counted. Approximately 64-89% of the cells in Scarpa's ganglion were labeled retrogradely following uvula-nodular injections. About 2% of the cells in the ipsilateral Scarpa's ganglion were labeled after injections of the flocculus. Virtually no cells were labeled following injections of the ventral paraflocculus. The vestibular primary afferent projection to the uvula-nodulus is so extensive that it must be part of a collateral system that also innervates the vestibular nuclei. This collateral projection pattern was confirmed by using fluorescent tracers injected into the uvula-nodulus and vestibular complex. Fluorogold was injected into the uvula-nodulus and peroxidase-rhodamine isothiocyanate was injected into the vestibular complex. More than 50% of the neurons in Scarpa's ganglion were double labeled by these subtotal injections. The dense vestibular primary afferent projection to the uvula-nodulus was confirmed by using the C fragment of tetanus toxin (TTC) injected into the labyrinth as an orthograde tracer. With the TTC technique, the vestibular primary afferent projection to the uvula-nodulus terminated exclusively in the ipsilateral granule cell layer of lobules 9d and 10. Much sparser vestibular primary afferent projections were found in the banks of major cerebellar sulci. A barely detectable projection was found to the flocculus and ventral paraflocculus.

The inferior olive is divided into several subnuclei that receive specific sensory information. The caudal dorsal cap of the medial accessory subdivision of the inferior olive receives horizontal optokinetic information from the nucleus of the optic tract. The immediately subjacent beta-nucleus receives vertical vestibular information mediated by a GABAergic pathway originating from the ipsilateral descending and medial vestibular nuclei. None of the transmitters to the dorsal cap have been identified. Using choline acetyltransferase (ChAT) immunohistochemistry, we have identified a cholinergic pathway that terminates exclusively in the dorsal cap of rats and monkeys. No other division of the inferior olive received a significant cholinergic innervation. In the rabbit, immunostaining for ChAT reveals a weaker and more diffuse cholinergic innervation of both the dorsal cap and the subjacent beta-nucleus. In rats and rabbits we injected iontophoretically the orthograde tracer Phaseolus vulgaris leucoagglutinin (PHA-L) into the medial and descending vestibular nuclei (MVN, DVN) as well as the nucleus prepositus hypoglossi (NPH) in order to trace the possible origin of the cholinergic projection. PHA-L injections into the NPH and medial aspect of the MVN labeled terminals within the contralateral dorsal cap. PHA-L injections in the central and lateral aspects of the MVN as well as the DVN labeled the ipsilateral beta-nucleus. Pressure injections of wheat germ agglutinin-horseradish peroxidase (WGA-HRP) in the caudal dorsal cap of the rabbit inferior olive demonstrated a predominantly contralateral projection to the dorsal cap from the lateral aspect of the NPH. However, pressure injections of HRP into the caudal dorsal cap combined with ChAT immunohistochemistry in the rabbit demonstrated that most of the neurons of the NPH that projected to the dorsal cap were not cholinergic, and that most of the ChAT-positive neurons within the NPH occupied a more ventral location than the neurons within the NPH that were retrogradely labeled from the HRP injection into the contralateral dorsal cap. In the rat, we made lesions in the MVN, DVN and NPH by injection of ibotenic acid (0.3-0.5 microliter), in an attempt to deplete the dorsal cap of the inferior olive of its cholinergic input. Lesions confined to the NPH and medial aspect of the MVN of the rat caused a loss of ChAT staining in the contralateral dorsal cap. Lesions placed more laterally within the MVN or DVN failed to deplete ChAT-positive terminals in the contralateral or ipsilateral dorsal caps. The dorsal cap of the rat and monkey receives a discrete cholinergic projection.(ABSTRACT TRUNCATED AT 400 WORDS)

Corticotropin-releasing factor (CRF) has been implicated by both anatomical and physiological techniques as a potential cerebellar transmitter or modulator. In the present experiment, with the aid of immunohistochemistry, we have described specific cerebellar afferent pathways in the rabbit in which CRF is located. CRF-immunoreactive climbing fibers were present in the molecular layer throughout the cerebellum, but especially in lobules 8-9a. All inferior olivary neurons were CRF-immunoreactive. In lobules 8-9a, CRF-immunoreactive mossy fibers were organized in sagittal bands. The highest density of CRF-immunoreactive mossy fiber terminals was observed in the granule cell layer of lobules 8-9a and the flocculus. No CRF-immunoreactive perikarya were located in rabbit cerebellum. The brainstem origin of CRF-immunoreactive mossy fiber terminals was suggested by numerous CRF-immunoreactive perikarya located in the medial, lateral and descending vestibular nuclei, nucleus prepositus hypoglossi, nucleus x, paramedian reticular nucleus, gigantocellular reticular nucleus, lateral reticular nucleus, and raphé nuclei. Using double label experiments, we investigated the specific CRF afferent projection to the flocculus and posterior vermis. Horseradish peroxidase (HRP) injections into the posterior vermis double labeled CRF-immunoreactive neurons in the caudal medial and descending vestibular nuclei and nucleus prepositus hypoglossi. HRP injections into the flocculus double labeled more CRF-immunoreactive neurons in the nucleus prepositus hypoglossi than in the vestibular nuclei. HRP injections into either the posterior vermis or flocculus double labeled CRF-immunoreactive neurons in the paramedian reticular nucleus, nucleus reticularis gigantocellularis, and raphé nuclei. These data suggest that CRF may play an important role in vestibularly related functions of the cerebellum.

. Cerebellar nodulectomy impairs spatial memory of vestibular and optokinetic stimulation in rabbits. J Neurophysiol 87: 962-975, 2002; 10.1152/jn.00528.2001. Natural vestibular and optokinetic stimulation were used to investigate the possible role of the cerebellar nodulus in the regulation and modification of reflexive eye movements in rabbits. The nodulus and folium 9d of the uvula were destroyed by surgical aspiration. Before and after nodulectomy the vertical and horizontal vestibuloocular reflexes (VVOR, HVOR) were measured during sinusoidal vestibular stimulation about the longitudinal (roll) and vertical (yaw) axes. Although the gain of the HVOR (G HVOR ϭ peak eye movement velocity/peak head velocity) was not affected by the nodulectomy, the gain of the VVOR (G VVOR ) was reduced. The gains of the vertical and horizontal optokinetic reflexes (G VOKR , G HOKR ) were measured during monocular, sinusoidal optokinetic stimulation (OKS) about the longitudinal and vertical axes. Following nodulectomy, there was no reduction in G VOKR or G HOKR . Long-term binocular OKS was used to generate optokinetic afternystagmus, OKAN II, that lasts for hours. After OKAN II was induced, rabbits were subjected to static pitch and roll, to determine how the plane and velocity of OKAN II is influenced by a changing vestibular environment. During static pitch, OKAN II slow phase remained aligned with earth-horizontal. This was true for normal and nodulectomized rabbits. During static roll, OKAN II remained aligned with earth-horizontal in normal rabbits. During static roll in nodulectomized rabbits, OKAN II slow phase developed a centripetal vertical drift. We examined the suppression and recovery of G VVOR following exposure to conflicting vertical OKS for 10 -30 min. This vestibular-optokinetic conflict reduced G VVOR in both normal and nodulectomized rabbits. The time course of recovery of G VVOR after conflicting OKS was the same before and after nodulectomy. In normal rabbits, the head pitch angle, at which peak OKAN II velocity occurred, corresponded to the head pitch angle maintained during long-term OKS. If the head was maintained in a "pitched-up" or "pitched-down" orientation during long-term OKS, the subsequently measured OKAN II peak velocity occurred at the same orientation. This was not true for nodulectomized rabbits, who had OKAN II peak velocities at head pitch angles independent of those maintained during long-term OKS. We conclude that the nodulus participates in the regulation of compensatory reflexive movements. The nodulus also influences "remembered" head position in space derived from previous optokinetic and vestibular stimulation.

Prolonged binocular optokinetic stimulation (OKS) in the rabbit induces a high-velocity negative optokinetic afternystagmus (OKAN II) that persists for several hours. We have taken advantage of this uniform nystagmus to study how changes in static head orientation in the pitch plane might influence the orientation of the nystagmus. After horizontal OKS, the rotation axis of the OKAN II remained almost constant in space as it was kept aligned with the gravity vector when the head was pitched by as much as 80 degrees up and 35 degrees down. Moreover, during reorientation, slow-phase eye velocity decreased according to the head pitch angle. Thereafter, we analyzed the space orientation of OKAN II after optokinetic stimulation during which the head and/or the OKS were pitched upward and downward. The rotation axis of OKAN II did not remain aligned with an earth vertical axis nor a head vertical axis, but it tended to be aligned with that of the OKS respace. The slow-phase eye velocity of OKAN II was also affected by the head pitch angle during OKS, because maximal OKAN II velocity occurred at the same head pitch angle as that during optokinetic stimulation. We suggest that OKAN II is coded in gravity-centered rather than in head-centered coordinates, but that this coordinate system may be influenced by optokinetic and vestibular stimulation. Moreover, the velocity attenuation of OKAN II seems to depend on the mismatch between the space-centered nystagmus rotation axis orientation and that of the "remembered" head-centered optokinetic pathway activated by OKS.

No abstract

The vestibulo-ocular reflexes (VORs) were studied in guinea pigs receiving daily administration of aminoglycoside antibiotics. The vestibular epithelia were also examined by scanning electron microscope technique (SEM). The treatment with aminoglycosides led to varying degrees of VORs according to (i) the type of aminoglycoside drug; (ii) the duration of the treatment and, (iii) the sensitivity of the various vestibular receptors. Gentamicin caused an earlier and severe reduction of the VOR gain. Dibekacin also caused evident damage, but the onset of its action was delayed. Both drugs affected mainly the vertical responses. Tobramycin and netilmicin altered the VORs slightly. Histological examination revealed damage to the sensory epithelia corresponding to the observed VOR impairments.

The contribution of the maculo-ocular reflex to gaze stability was studied in 10 pigmented rabbits by rolling the animals at various angles of sagittal inclination of the rotation and/or longitudinal animal axes. At low frequencies (0.005-0.01 Hz) of sinusoidal stimulation the vestibulo-ocular reflex (VOR) was due to macular activation, while at intermediate and high frequencies it was mainly due to ampullar activation. The following results were obtained: 1) maculo-ocular reflex gain decreased as a function of the cosine of the angle between the rotation axis and the earth's horizontal plane. No change in gain was observed when longitudinal animal axis alone was inclined. 2) At 0 degrees of rotation axis and with the animal's longitudinal axis inclination also set at 0 degrees, the maculo-ocular reflex was oriented about 20 degrees forward and upward with respect to the earth's vertical axis. This orientation remained constant with sagittal inclinations of the rotation and/or longitudinal animal axes ranging from approximately 5 degrees upward to 30 degrees downward. When the longitudinal animal axis was inclined beyond these limits, the eye trajectory tended to follow the axis inclination. In the upside down position, the maculo-ocular reflex was anticompensatory, oblique and fixed with respect to orbital coordinates. 3) Ampullo-ocular reflex gain did not change with inclinations of the rotation and/or longitudinal animal axes. The ocular responses were consistently oriented to the stimulus plane. At intermediate frequencies the eye movement trajectory was elliptic because of directional differences between the ampullo- and maculo-ocular reflexes.(ABSTRACT TRUNCATED AT 250 WORDS)