Anterior cingulate neurons display subregion-specific interaction with frontal eye fields as revealed by combined antidromic stimulation and resting state imaging

The anterior cingulate cortex (ACC) is thought to exert cognitive control over saccade generation in the frontal eye fields (FEF), but the nature of this interaction remains unclear. Although prior imaging studies have suggested ACC interacts with FEF, few studies have confirmed this by electrophysiological recordings. This study aimed to characterize the functional connectivity between ACC and medial and lateral FEF during cognitive saccade tasks. We combined resting-state functional MRI (rs-fMRI) with single-unit electrophysiology in two macaque monkeys performing memory-guided saccade and pro-/anti-saccade tasks. Anti- and ortho-dromic stimulation was used to electrophysiologically identify ACC neurons mono- and polysynaptically connected to FEF. We analyzed ACC neuronal selectivity for different task aspects and correlated these properties with both positive and negative rs-fMRI functional connectivity between ACC and FEF subregions. Anti- and ortho-dromically identified ACC neurons were predominantly connected to medial FEF, which showed stronger positive functional connectivity with ACC compared to lateral FEF. Sites with higher proportions of task-selective neurons yielded stronger functional connectivity with FEF. This stronger functional connectivity was particularly related to the delay and saccadic periods of different cognitive saccade tasks. Using combined imaging and electrophysiology, our findings provide converging evidence for functional interactions between ACC and FEF, predominantly medial FEF regions which encode large amplitude saccades. The correlation between functional connectivity and task-related neuronal selectivity supports ACCs interaction with FEF in the modulation of saccade generation and cognitive control. Additionally, we report suggestive evidence that mono- and poly-synaptic connections may be related to positive functional connectivity, but we found no such relationship for negative functional connectivity (anticorrelations). These results advance our understanding of prefrontal cortical interactions in oculomotor control and the electrophysiological mechanisms of positive and negative resting-state functional connectivity.