Perturbation of combined saccade-vergence movements by microstimulation inmonkey superior colliculus

This study investigated the role of the monkey superior colliculus (SC) in the control of visually (V)-guided combined saccade-vergence movements by a ssessing the perturbing effects of microstimulation. We elicited an electri cal saccade (E) by stimulation (in 20% of trials) in the SC while the monke y was preparing a V-guided movement to a near target. The target was aligne d such that E- and V-induced saccades had similar amplitudes but different directions and such that V-induced saccades had a significant vergence comp onent (saccades to a near target). The onset of the E-stimulus was varied f rom immediately after V-target onset to after V-saccade onset. E-control tr ials, where stimulation was applied during fixation of a V-target, yielded the expected saccade but no vergence. By contrast, early perturbation trial s, where the E-stimulus was applied soon after the onset of the V-target, c aused an E-triggered response with a clear vergence component toward the V- target. Midflight perturbation, timed to occur just after the monkey initia ted the movement toward the target, markedly curtailed the ongoing vergence component during the saccade. Examination of pooled responses from both ty pes of perturbation trials showed weighted-averaging effects between E- and V-stimuli in both saccade and fast vergence components. Both components ex hibited a progression from E- to V-dominance as the E-stimulus was delayed further. This study shows that artificial intervention in the SC, while a t hree-dimensional (3D) refixation is being prepared or is ongoing, can affec t the timing (WHEN) and the metric specification (WHERE) Of both saccades a nd vergence. To explain this we interpret the absence of overt vergence in the E-controls as being caused by a zero-vergence change command rather tha n reflecting the mere absence of a collicular vergence signal. In the pertu rbation trials, the E-evoked zero-vergence signal competes with the V-initi ated saccade-vergence signal, thereby,giving rise to a compromised 3D respo nse. This effect would be expected if the population of movement cells at e ach SC site is tuned in 3D, combining the well-known topographical code for direction and amplitude with a nontopographical depth representation. On E -stimulation, the local population would yield a net saccade signal caused by the topography, but the cells coding for different depths would be excit ed equally, causing the vergence change to be zero.