Our cognitive abilities rely on dynamic interactions of neuronal groups distributed across distant cortical areas. How the efficiency of these interactions is modulated flexibly is an unresolved question in neuroscience. Previous studies suggest that the efficacy of mutual inputs between two neuronal groups is shaped by their rhythmic activities. In particular, neuronal groups that synchronize their rhythmic activity could establish a communication structure that enhances their interaction dynamically, while reducing the influence from non-synchronized neuronal groups.
Here, we tested this hypothesis of communication through coherence directly by testing the strength of intra- and inter-areal interactions as a function of the precision of their synchronization. We specifically hypothesized that interactions between distant cortical neuronal groups should be enhanced during periods of precise synchronization.
We tested this by recording spike- and local-field potential (LFP) activity simultaneously from electrodes in areas 17, 18 and 21a of awake cats, stimulated with moving gratings. Each electrode assessed the activation in a local group of neurons. While stimulation was constant, the spontaneous response variability allowed our analysis. We investigated whether the precision of synchronization (within a given trial) between neuronal groups predicted the strength of their interaction. We estimated interaction strength by determining the Spearman rank correlation coefficient (across trials) between their firing rates, or between the spectral power of their spike trains and their LFPs.
We found that neuronal groups interacted stronger when they were precisely synchronized. Both the precision of synchronization and the strength of interaction fluctuated rapidly in time and precise synchronization preceded strong interaction by a few milliseconds, suggesting a causal role. Synchronization-dependent interactions were present between neuronal groups both within the same area and distributed in separate areas. Within an area, the effect was present for distances between 2 and 15 mm.
These findings suggest that intra-areal and inter-areal neuronal
interactions are mechanistically subserved by precise neuronal
synchronization. We hypothesize that neuronal synchronization establishes
the dynamic and specific patterns of neuronal interactions that are at the
core of our cognitive abilities.