Non-invasive assessment of superficial and deep layer circuits in human motor cortex

Key points: The first indirect (I) corticospinal volley from stimulation of the motor cortex consists of two parts: one that originates from infragranular layer 5 and a subsequent part with a delay of 0.6 ms to which supragranular layers contribute. Non-invasive probing of these two parts was performed in humans using a refined electrophysiological method involving transcranial magnetic stimulation and peripheral nerve stimulation. Activity modulation of these two parts during a sensorimotor discrimination task was consistent with previous results in monkeys obtained with laminar recordings.

Abstract: Circuits in superficial and deep layers play distinct roles in cortical computation, but current methods to study them in humans are limited. Here, we developed a novel approach for non-invasive assessment of layer-specific activity in the human motor cortex. We first conducted brain slice and in vivo experiments on monkey motor cortex to investigate the output timing from layer 5 (including corticospinal neurons) following extracellular stimulation. Neuron responses contained cyclical waves. The first wave was composed of two parts: the earliest part originated only from stimulation of layer 5; after 0.6 ms, stimuli to superficial layers 2/3 could also contribute. In healthy humans we then assessed different parts of the first corticospinal volley elicited by transcranial magnetic stimulation (TMS), by interacting TMS with stimulation of the median nerve generating an H-reflex. By adjusting the delay between stimuli, we could assess the earliest volley evoked by TMS, and the part 0.6 ms later. Measurements were made while subjects performed a visuo-motor discrimination task, which has been previously shown in monkey to modulate superficial motor cortical cells selectively depending on task difficulty. We showed a similar selective modulation of the later part of the TMS volley, as expected if this part of the volley is sensitive to superficial cortical excitability. We conclude that it is possible to segregate different cortical circuits which may refer to different motor cortex layers in humans, by exploiting small time differences in the corticospinal volleys evoked by non-invasive stimulation.