Dynamics of adaptive change in vestibulo-ocular reflex direction. I. Rotations in the horizontal plane.

Horizontal and vertical vestibulo-ocular reflex (VOR) eye movements were recorded in alert cats before and after adaptation to vertical optokinetic motion coupled with horizontal rotation at 0.05, 0.25 or 1.0 Hz. Within 15-30 min, the VOR measured in darkness acquired a vertical component; the maximal directional change in the VOR occurred at the frequency of the adapting stimulus. At other frequencies the gain was less and there were phase leads or lags for higher or lower frequencies, respectively. Adaptive VOR was stable for at least 14 h in unrestrained animals with no visual input and decayed within 30 min during rotation in a stationary visual world.

Phase-shifting effects of acute increases in activity on circadian locomotor rhythms in hamsters.

Experiments were conducted in golden hamsters to examine the relationship between induced acute increases in locomotor activity and phase shifts in the circadian clock underlying the rhythm of activity. Injections of the short-acting benzodiazepine triazolam (TZ) 6 h before the onset of activity resulted in an acute increase in activity and a phase advance in the rhythm of activity; injections of TZ induced larger phase shifts in animals housed without running wheels than in those housed with wheels. Transfer to a cage with access to a running wheel for 1 h at different circadian times induced large phase advances (mean of 2 h) and small phase delays depending on the circadian time of transfer. Maximal mean phase advances resulted when animals were transferred to a cage with wheel 3 h before activity onset, and at this circadian time there was a significant correlation between the magnitude of the phase shift and the amount of increase over baseline activity for the first hour after transfer. These results indicate that access to a running wheel in animals housed without wheels can be a significant phase-shifting stimulus to the circadian clock and that the phase shifts induced by injection of TZ or transfer to a new cage with wheel are related to the activity state of the animal or to the amount of locomotor activity that is induced at particular times.

Torque vectors of neck muscles in the cat.

Anatomical texts describe the neck musculature without measurements of muscle locations or quantitative estimates of pulling actions (torques). This study is based on measurements in stereotaxic coordinates of cat neck muscle origins and insertions, and neck intervertebral rotation axes. Torque vectors in three dimensions were calculated for 14 pairs of dorsal and ventral muscles that insert on the skull or first cervical vertebra. Predicted torque vectors were in general agreement with qualitative statements in the literature. Biventer cervicis and the rectus capitis major, medius, and minor muscles act mainly to raise the head, and longus capitis acts almost exclusively to lower the head. Longissimus capitis, sternomastoid, and cleidomastoid act mainly to roll the head. Complexus acts about equally to raise the head and rool it. Splenius and occipitoscapularis have torque in all three coordinate directions. Torques were altered by changing the pitch of the head with respect to the neck. The calculated neck muscle torques did not correspond to previously reported directions of neck muscle excitation during the vestibulocollic reflex. The neck musculature appears to be a complex, multidimensional system that presents interesting problems in motor control.

Models of sensorimotor transformations and vestibular reflexes.

The vestibulo-ocular and vestibulo-collic reflexes are well-studied sensorimotor systems with dynamic properties that have been successfully modeled. Recently proposed matrix and tensorial models attempt to describe the spatial organization of these reflexes in three dimensions. Here we describe experiments that test these models. We show that a matrix model of the vestibulo-ocular reflex provides a satisfactory description of its spatial properties. The vestibulo-collic reflex is more complex, but a tensorial model makes close predictions of neck muscle excitation by the vestibulo-collic reflex. In addition, our preliminary data show that the cervico-collic or neck stretch reflex produces essentially the same spatial pattern of neck muscle excitation as the vestibulo-collic reflex, a finding predicted by the tensorial model. We conclude by showing electromyographic and single neuron responses that can be modeled only by combining models of dynamics with models of spatial organization. We believe that the development of such models is the next major challenge in the application of quantitative methods to analysis of reflex behavior.