Gaze position effects and position-dependent motor tuning in primate superior colliculus (SC) neurons during head-unrestrained visually guided movements

J. F. DESOUZA, *X. YAN, G. BLOHM, H. WANG, J. D. CRAWFORD
Ctr Vision Res, York Univ, Toronto, ON, CANADA.

Neural activity in the SC is highly correlated with gaze shifts composed of both eye and head movements (Freedman and Sparks, 1996; Munoz et al. 1991). Further, SC stimulation evokes eye and head gaze shifts that converge as a function of amplitude and initial gaze position, consistent with an eye-fixed motor code for gaze (Klier et al. 2001). We hypothesize that SC neural activity correlates best with gaze target location in retinal coordinates. This predicts that the optimal directional tuning of SC neurons will change as a specific function of amplitude tuning and initial gaze position (Smith and Crawford 2005). Electrical stimulation and/or visual receptive field examination are being used to estimate the optimal gaze amplitude and direction for each recording location. Monkeys randomly fixate one of three different initial gaze directions each separated by 20+/-10 degrees for 500 ms and then make their head-free gaze shift to one of five visual targets placed along a semi-circle of iso-amplitude targets (centered around the position of the receptive field maximum for straight-ahead gaze). To date, we have recorded from thirty-four SC neurons that have been fully tested in the head-unrestrained paradigm. Analysis of these neurons show that SC neurons do indeed show strong initial gaze position dependent firing changes during head-unrestrained gaze shifts. Furthermore, a large majority of our SC neurons show a position-dependent modulation of their firing rates as a function of the initial gaze position, as required for a non-linear transformation of retinal coordinates into gaze motor coordinates (Smith and Crawford 2005).