Saccades to somatosensory targets. III. eye-position-dependent somatosensory activity in primate superior colliculus

1. We recorded from cells with sensory responses to somatosensory stimuli in the superior colliculus (SC) of awake monkeys. Our goal was to determine the frame of reference of collicular somatosensory signals by seeing whether the positions of the eyes influenced the responses of cells to a given tactile stimulus. Somatosensory targets consisted of vibrotactile stimuli delivered to the hands, which were held in fixed spatial positions. Monkeys performed a delayed saccade task from different initial fixation positions to the locations of these tactile stimuli or to visual stimuli at approximately the same location. 2. The responses of a majority of somatosensory cells (25 of 34 or 74%) were significantly affected by eye position. Nearly all somatosensory cells also responded to visual targets (28 of 30, 93%). Cells whose somatosensory responses depended on eye position responded to visual and somatosensory targets located at approximately the same direction in space with respect to the eyes. 3. The activity of these cells exhibited both sensory and motor qualities. The discharge was more closely linked in time to stimulus onset than to the movement. Sensory features of the stimulus were reflected in the responses: the discharge of a number of cells was phase-locked to the pulses of vibration. The sensory responses occurred even if the animal's next saccade was not directed into the response field of the cell. However, two thirds of the cells also exhibited a burst of motor activity in conjunction with the saccade to the somatosensory target. Sensory and motor activity were not always spatially coextensive. When different, the tuning of motor activity was broader. 4. Cells with somatosensory responses to vibratory stimulation of the hands were found in a wide region of the SC, spanning a 40 degrees range of movement amplitudes. 5. These data show that somatosensory signals in the SC are not purely somatotopic but are dependent on eye position. For stimuli at a fixed location, this eye position dependence allows somatosensory and visual signals to be in register and share a premotor circuitry for guiding saccadic eye movements. 6. The dependence of the somatosensory responses on eye position suggests that the somatosensory receptive fields may either shift on the body surface or they may be restricted to a limited region of the body surface but be gated by eye (and body) position. Future experiments varying body position and the location of the stimulus on the body surface are needed to determine which of these alternatives is correct. Cells with either type of receptive field could provide an unambiguous signal of the location of somatosensory saccade targets with respect to the eyes. The transformation of somatosensory signals from a body-centered frame of reference to a frame of reference that depends on the position of the stimulus with respect to the eyes is necessary for the correct activation of collicular neurons with motor activity, because this activity encodes saccades as desired changes in eye position.