Attention induces dynamic changes in coherence between monkey area V1 and V4

Neurons receive a vast number of synaptic inputs at the same time. To successfully process a relevant piece of information, a neuron needs to select relevant, while suppressing irrelevant signals. In order to dynamically route a meaningful set of signals through processing pathways, neurons need to have a fast and flexible mechanism to change their effective connectivity. Here, we hypothesize that attention changes effective connectivity by modulating the synchrony between neuronal populations.We investigated whether attention dynamically changes effective connectivity between area V4 neurons and their V1 input. Single unit activity, multi unit activity and local field potentials (LFP) were recorded simultaneously from macaque monkeys’ areas V1 and V4. While maintaining their gaze on a fixation spot, the monkeys had to attend to one of two continuously morphing shapes and respond to reoccurrence of the initial sample shape in the attended stream. Size and position of the shapes were adjusted such that both fitted into a single V4 receptive field (RF) while covering two separate, non-overlapping V1 RFs.We found strong phase-coherence of gamma-band LFPs between the V1 populations responding to the attended stimulus and the V4 population. At the same time, V1 populations responding to the unattended stimulus showed only weak LFP phase-coherence with the V4 population. Furthermore, spike-triggered averages (STAs) of at least 40% of the recording pairs revealed that spikes recorded from area V4 were precisely timed with respect to the phase of the gamma oscillation in the upstream area V1. More importantly, almost all these cases showed an increased effect with attention. These results support the hypothesis that attention induces dynamic changes in the effective connectivity between two anatomically hard-wired neuronal populations by selectively modulating synchrony.