Visual-vestibular responses in vestibular nuclear neurons in the intact and cerebellectomized, alert cat

Abstract

In the alert cat a very large majority of vestibular nucleus neurons related specifically to the horizontal semicircular canals responded equally well to rotations of the visual surroundings. These visual responses in vestibular neurons remained qualitatively unchanged in cerebellectomized cats. Units in paralyzed animals showed a response that was proportional to ‘surround velocity’ only over the lowest portion (3–10 deg./s) of the range of stimulus velocities tested. The response then saturated, or declined, but some response was elicited even at the highest tested velocity (60 deg./s). The velocity selectivity of the response in cerebellectomized cats was similar, but it became saturated at lower stimulus velocities and was even less at higher velocites (30 and 60 deg./s). Sinusoidal analysis of the visual response of vestibular neurons indicated that two types of units exist. The low-frequency type (a majority of the units) showed a response that already lagged the ‘surround velocity’ at 0.025 Hz by about 50 deg. The other type of vestibular neuron showed a surprisingly well maintained response with only a moderate phase lag even at 0.5 Hz. Computer analysis and vector addition of the separate responses to pure vestibular and pure visual stimulation led to the predicted responses to combined visual-vestibular stimulation that compared very closely to the combined stimulus responses that were actually measured, in both phase and gain. In concordance with the unit data, measured eye movements continued to show many of the well-known influences of visual input in the cerebellectomized animals. The low-frequency gain and phase improvement of the vestibulo-ocular reflex in the presence of visual input was still present. Postrotatory nystagmus was still suppressed more quickly in the presence of vision. Optokinetic eye movements were normal at low surround velocity (< 10 deg./s), but were extremely velocity-limited at higher stimulus speeds. The cat therefore is similar to other tested vertebrate species in showing visual influences on vestibular neurons. Contrary to prevailing hypotheses, the cerebellum, and the flocculus specifically, are not essential parts of the pathway mediating these visual influences. Sinusoidal analysis suggests that in the cat the combination of visual and Vestibular inputs at the level of the vestibular nucleus is essentially linear. At least some vestibular neurons continue to respond to higher frequency visual input, suggesting a role for visual input in visual-vestibular interaction even at high frequencies.