SPATIAL COORDINATION BY DESCENDING VESTIBULAR SIGNALS .2. RESPONSE PROPERTIES OF MEDIAL AND LATERAL VESTIBULOSPINAL TRACT NEURONS IN ALERT AND DECEREBRATE CATS

Spatial response properties of medial (MVST) and lateral (LVST) vestib ulospinal tract neurons were studied in alert and decerebrate cats dur ing sinusoidal angular rotations of the whole body in the horizontal a nd many vertical planes. Of 220 vestibulospinal neurons with activity modulated during 0.5-Hz sinusoidal rotations, 200 neurons exhibited re sponse gains that varied as a cosine function of stimulus orientation and phases that were near head velocity for rotation planes far from t he minimum response plane. A maximum activation direction vector (MAD) , which represents the axis and direction of rotation that maximally e xcites the neuron, was calculated for these neurons. Spatial propertie s of secondary MVST neurons in alert and decerebrate animals were simi lar. The responses of 88 of 134 neurons (66%) could be accounted for b y input from one semicircular canal pair. Of these, 84 had responses c onsistent with excitation from the ipsilateral canal of the pair (13 h orizontal, 27 anterior, 44 posterior) and 4 with excitation from the c ontralateral horizontal canal. The responses of the remaining 46 (34%) neurons suggested convergent inputs. The activity of 38 of these was significantly modulated by both horizontal and vertical rotations. Twe lve neurons (9%) had responses that were consistent with input from bo th vertical canal pairs, including 9 cells with MADs near the roll axi s. Thirty-two secondary MVST neurons (24%) had type II yaw and/or roll responses. The spatial response properties of 18 secondary LVST neuro ns, all studied in decerebrate animals, were different from those of s econdary MVST neurons. Sixteen neurons (89%) had type II yaw and/or ro ll responses, and 12 (67%) appeared to receive convergent canal pair i nput. Convergent input was more common on higher-order vestibulospinal neurons than on secondary neurons. These results suggest that MVST an d LVST neurons and previously reported vestibule-ocular neurons transm it functionally different signals. LVST neurons, particularly those wi th MADs close to the roll axis, may be involved in the vestibular-limb reflex. The combination of vertical and ipsilateral horizontal canal input on many secondary MVST neurons suggests a contribution to the ve stibulocollic reflex. However, in contrast to most neck muscles, very few neurons had maximum vertical responses near pitch.