Distributed representations of the "preparatory set" in the frontal oculomotor system: a TMS study

The generation of saccades is influenced by the level of "preparatory set activity" in cortical oculomotor areas. This preparatory activity can be examined using the gap-paradigm in which a temporal gap is introduced between the disappearance of a central fixation target and the appearance of an eccentric target.

Ten healthy subjects made horizontal pro- or antisaccades in response to lateralized cues after a gap period of 200 ms. Single-pulse transcranial magnetic stimulation (TMS) was applied to the dorsolateral prefrontal cortex (DLPFC), frontal eye field (FEF), or supplementary eye field (SEF) of the right hemisphere 100 or 200 ms after the disappearance of the fixation point. Saccade latencies were measured to probe the disruptive effect of TMS on saccade preparation. In six individuals, we gave realistic sham TMS during the gap period to mimic auditory and somatosensory stimulation without stimulating the cortex.

TMS to DLPFC, FEF, or SEF increased the latencies of contraversive pro- and antisaccades. This TMS-induced delay of saccade initiation was particularly evident in conditions with a relatively high level of preparatory set activity: The increase in saccade latency was more pronounced at the end of the gap period and when participants prepared for prosaccades rather than antisaccades. Although the "lesion effect" of TMS was stronger with prefrontal TMS, TMS to FEF or SEF also interfered with the initiation of saccades. The delay in saccade onset induced by real TMS was not caused by non-specific effects because sham stimulation shortened the latencies of contra- and ipsiversive anti-saccades, presumably due to intersensory facilitation.

Our results are compatible with the view that the "preparatory set" for contraversive saccades is represented in a distributed cortical network, including the contralateral DLPFC, FEF and SEF.