Characterization of resting state dynamics in monkey motor cortex

Nowadays, modeling studies of cortical network dynamics aim to include realistic assumptions on the neuronal properties [1,2]. However, such models do not consider functional aspects – they rather describe the “ground” or “resting state” of cortical networks, typically characterized as asynchronous irregular spiking [2]. For model validation, i.e., for a concrete comparison of experimental versus model data, one needs experimental “resting state” data. Therefore we performed a “resting state” experiment (a term adapted from human fMRI studies, defined as brain activity observed when the subject is “at rest” without any task or particular stimulation [3]), which is in contrast to most neurophysiological studies [e.g. 4]. We recorded the neuronal activity from macaque monkey motor cortex with a chronically implanted 4x4mm2 100 electrode Utah Array (Blackrock Microsystems), while the monkey was sitting in a chair “resting”, for 15min. Based on a video recording of this experiment, we differentiate between “resting” intervals and intervals when the monkey spontaneously moved.The goal of this study is to thoroughly characterize the simultaneous spiking activity recorded from 146 single units during resting state. To enable a detailed comparison to simulated spiking data, we subdivide the single units into putative excitatory and inhibitory neurons based on their spike shapes [5]. We apply common statistical measures, e.g., firing rate, (local) coefficient of variation for single unit characterization, and we also compute the pairwise fine temporal correlation by correlation coefficients. Comparing the distributions of these measures we find hardly any difference between our two behavioral states. However, when focusing on single units we notice that some neurons increase their firing rates systematically when the monkey moves compared to rest, whereas others decrease or do not change their rates. Thus, there was seemingly no difference on the population level, but significant differences on the level of individual neurons. Moreover, we observe a strong temporal correlation between a few neuronal units, independent of their cortical distance, while others show lower or no correlation. Our next steps are to characterize if such findings are particularly different for excitatory and inhibitory neurons. Further, we aim to study the underlying network mechanisms. One possibility would be to re-consider the balancing effects of inhibition and recurrence [5,6]