Responses to head tilt were recorded from vestibular neurons in and around the lateral vestibular nucleus (LVN) of the decerebrate cat. Each animal had all six semicircular canals rendered nonfunctional by a plugging procedure. Each cell was studied by slowly tilting the cat, using one or both of two paradigms. In the first method, sinusoidal tilts (0.05 or 0.1 Hz) were used to produce bidirectional stimuli in up to 12 pairs of directions, including left/right (roll tilt) and fore/aft (pitch). The second method imposed a constant 10 degree tilt; the direction of the tilt was rotated around the animal by an appropriate combination of roll and pitch motions. Neurons responded by maximally increasing their discharge frequency in a particular direction of head tilt from the horizontal. Each cell's response could be described by a vector in the animal's horizontal plane whose orientation is given by the direction of the most effective stimulus and whose length represents the neuron's maximal sensitivity to tilt. The two methods of stimulation yielded equivalent response vectors. Response vectors were obtained for 100 neurons. The distribution of vector directions for these vestibular neurons was not uniform; there was a conspicuous absence of neurons with fore/aft-directed vectors. The sensitivity of these cells (length of the response vector) ranged from 10 to 230 impulses X s-1 X g-1 (median 50). Neurons whose vectors lay in the ipsilateral half-plane (which would be excited by ear-down tilt) tended to be less sensitive than those with contralateral vectors. Neurons excited by ear-up tilt tended to be located ventrally in the LVN, while those excited by ear-down tilt were more evenly distributed. There was no other obvious correlation of vector orientation with the anatomical locus of the cell in the LVN. The directional selectivity of the responses of these neurons to head tilts are similar to those previously reported tin utricular afferents. The broad distribution of response vector orientations provides an appropriate substrate for vestibulospinal reflexes to a wide variety of head tilts.

1. We have studied, in decerebrate cats, the responses of neurons in the lateral and descending vestibular nuclei to whole-body rotations in vertical planes that activated vertical semicircular canal and utricular receptors. Some neurons were identified as vestibulospinal by antidromic stimulation with floating electrodes placed in C4. 2. The direction of tilt that caused maximal excitation (response vector orientation) of each neuron was determined. Neuron dynamics were then studied with sinusoidal stimuli closely aligned with the response vector orientation, in the range 0.02-1 Hz. A few cells, for which we could not identify a response vector, probably had spatial-temporal convergence. 3. On the basis of dynamics, neurons were classified as receiving their input primarily from vertical semicircular canals, primarily from the otolith organs, or from canal+otolith convergence. 4. Response vector orientations of canal-driven neurons were often near +45 degrees or -45 degrees with respect to the transverse (roll) plane, suggesting these neurons received excitatory input from the ipsilateral anterior or posterior canal, respectively. Some neurons had canal-related dynamics but vector orientations near roll, presumably because they received convergent input from the ipsilateral anterior and posterior canals. Few neurons had their vectors near pitch. 5. In the lateral vestibular nucleus, neurons with otolith organ input (pure otolith or otolith+canal) tended to have vector orientations closer to roll than to pitch. In the descending nucleus the responses were evenly divided between the roll and pitch quadrants. 6. We conclude that most of our neurons have dynamics and response vector orientations that make them good candidates to participate in vestibulospinal reflexes acting on the limbs, but not those acting on the neck.

1. To compare the properties of the vestibulocollic reflex (VCR) with those of vestibular neurons projecting to the neck [vestibulocollic (VC) neurons], we have studied the behavior of the latter in the decerebrate cat. Neurons were identified by their antidromic responses to stimulation in C1-C2, but not C5. Responses to stimulation of vestibular and neck receptors were produced by rotation of the body and head in vertical planes. 2. We determined the plane of whole body (vestibular) or body with head counter-rotated (neck) rotation, which produced the maximal modulation of each neuron (response vector orientation). Neuron dynamics were then studied with sinusoidal (0.02-2 Hz) stimuli aligned with this orientation. 3. On the basis of dynamics and vector orientation, the neuron was assigned a vestibular input classification of otolith, vertical canal, otolith + canal, or spatial-temporal convergence (STC). 4. The properties of this sample of VC neurons are similar to those of a larger population of vestibular neurons whose projection was not identified. For example, the distributions of cells with different types of vestibular inputs were roughly the same; in particular, few cells showed STC responses. In addition, there was no evidence of significant convergence of like canals across the midline (e.g., right anterior + left anterior). 5. Also similar to the larger unidentified population, 80% of VC neurons tested for neck input received such an input. The neck and vestibular responses tended to be antagonistic; the vector orientations were usually opposite, and the response gains and phases similar.(ABSTRACT TRUNCATED AT 250 WORDS)

The purpose of this review is to assess the role of short-latency pathways in the vestibulocollic reflex (VCR). First the current knowledge about the disynaptic and trisynaptic pathways linking semicircular canal and otolith afferents with cat neck motoneurons is summarized. We then discuss whether these pathways are sufficient or necessary to produce the responses observed in neck muscles by natural vestibular stimulation and conclude that they are neither. Finally, alternate pathways are considered, most likely involving reticulospinal fibers, which are an important part of the neural substrate of the VCR.

Although many previous experiments have considered the responses of vestibular nucleus neurons to rotations and translations of the head, little data are available regarding cells in the caudalmost portions of the vestibular nuclei (CVN), which mediate vestibulo-autonomic responses among other functions. This study examined the responses of CVN neurons of conscious cats to rotations in vertical planes, both before and after a bilateral vestibular neurectomy. None of the units included in the data sample had eye movement-related activity. In labyrinth-intact animals, some CVN neurons (22%) exhibited graviceptive responses consistent with inputs from otolith organs, but most (55%) had dynamic responses with phases synchronized with stimulus velocity. Furthermore, the large majority of CVN neurons had response vector orientations that were aligned either near the roll or vertical canal planes, and only 18% of cells were preferentially activated by pitch rotations. Sustained head-up rotations of the body provide challenges to the cardiovascular system and breathing, and thus the response dynamics of the large majority of CVN neurons were dissimilar to those of posturally-related autonomic reflexes. These data suggest that vestibular influences on autonomic control mediated by the CVN are more complex than previously envisioned, and likely involve considerable processing and integration of signals by brainstem regions involved in cardiovascular and respiratory regulation. Following a bilateral vestibular neurectomy, CVN neurons regained spontaneous activity within 24 h, and a very few neurons (<10%) responded to vertical tilts <15° in amplitude. These findings indicate that nonlabyrinthine inputs are likely important in sustaining the activity of CVN neurons; thus, these inputs may play a role in functional recovery following peripheral vestibular lesions.