To stay or to go? Paired associative stimulation highlights state-dependent causal interactions from right IFG to left M1 during a Go/NoGo task

Motor inhibition requires suppressing inappropriate motor responses. Studies suggest that during motor inhibition tasks (e.g., Go/NoGo Task) the right inferior frontal gyrus (rIFG) exerts an indirect influence over the primary motor cortex (M1) via a cortico-subcortical route encompassing the basal ganglia. Here, we sought to investigate whether strengthening a direct cortico-cortical rIFG-to-M1 route affects behavioural and electrophysiological indices of motor inhibition. We used transcranial magnetic stimulation (TMS) to record motor-evoked potentials (MEPs) to left M1 (lM1) stimulation; and induce associative plasticity of the rIFG-lM1 pathway using cortico-cortical Paired Associative Stimulation (ccPAS). Motor inhibition was assessed with a Go/NoGo task. In 2 days/sessions, 28 healthy humans received a ccPAS protocol aimed either at strengthening (ccPASrIFG-M1) the rIFG-to-lM1 pathway and a control protocol (ccPASM1-rIFG). In each session, participants performed a simple Go/NoGo task before, immediately after and 30 minutes after the ccPAS. MEPs to single- (M1) and dual-coil TMS (rIFG-M1) during observation of Go and NoGo stimuli were recorded before and after ccPAS. Participants were divided into two groups: ccPAS-responding, i.e., displaying “online” MEP changes already during ccPASrIFG-M1 administration, and non ccPAS-responding. Both groups showed similar increase of commission errors, reduction of omission errors, and reduced response times, irrespective of the ccPAS session. Concerning the neurophysiological results, ccPASrIFG-M1 induced distinct changes of rIFG-lM1 interactions during Go and NoGo trials: relative to single-pulse MEPs, dual-coil MEPs were enhanced in Go trials and suppressed in NoGo trials. This effect was significant only in ccPAS-responding participants. The ccPASM1-rIFG protocol did not affect MEP amplitudes. These findings show that enhancing rIFG-to-lM1 connectivity in ccPAS-responding participants, results in an enhanced state-dependent causal influence of rIFG over lM1, dynamically shifting from facilitatory to inhibitory depending on task demands. Therefore, while prior work has shown that the cortico-subcortical rIFG-to-lM1 route is selective for motor inhibition, here we show that the cortico-cortical rIFG-to-lM1 route is functionally malleable by ccPAS, subject to interindividual differences in sensitivity to ccPAS and flexibly modulated according to learned visuo-motor associations. By driving associative plasticity into the rIFG-to-lM1 network, we induced task-dependent modulations of rIFG-lM1 interactions only in ccPAS-responding participants, calling for the need of personalized brain stimulation approaches: future research should elucidate which factors contribute to interindividual variability and how to optimize stimulation protocols. Finally, we provide novel insights into the malleability and modulatory role of the cortico-cortical route connecting rIFG to lM1 as a flexible state-dependent neural circuit.