Investigating cortical oscillations in a data-driven multiscale model of macaque auditory thalamocortical circuits

Abstract

Cortical oscillations play a crucial role in processing rhythmic auditory stimuli such as speech. Here we developed a model of the macaque auditory pathway in order to investigate the mechanisms underlying cortical oscillations and their entrainment to rhythmic, low-frequency auditory stimuli. The simulated pathway extended from cochlea to cortex and included a biophysically-detailed, multiscale model of the auditory thalamocortical circuit, with medial geniculate body, thalamic reticular nucleus, and a column of primary auditory cortex (A1). Electrophysiology signals such as local field potential (LFP), current source density (CSD), and electroencephalography (EEG) data were recorded in silico from the structures in the modeled thalamocortical circuit. These data revealed physiologically realistic cortical oscillations that emerged spontaneously in the model. The full repertoire of frequency band oscillations was simulated in silico, and were comparable to those recorded in vivo from macaque non-human primate (NHP) A1. The biological detail of the model also allowed us to examine the contribution of individual cell-type and layer-specific neuron populations to specific oscillation events. We also demonstrated that the model undergoes physiologically-realistic phase entrainment to low- frequency rhythmic auditory stimuli. Deficits in low-frequency phase entrainment are often noted in schizophrenia, and may underlie some of the speech and auditory processing impairments seen in this disorder. In the model, we were able to induce a low-frequency entrainment deficit by simulating a subanesthetic dose of ketamine and reducing the strength of NMDA synapses onto cortical inhibitory interneurons. This in silico intervention thus allowed us to make a circuit-level prediction regarding the mechanisms underlying low-frequency entrainment deficits in schizophrenia. This model used the NEURON simulator and NetPyNE modeling tool to integrate information at the subcellular, cellular, and circuit-level scales, from synapse characteristics to cell electrophysiology to long- range, local and dendritic connectivity. Macaques were used as the preferred animal model for this study, due to similarities in the hierarchical structure of oscillatory activity in humans and non-human primates. This model thus incorporated multiscale information with macaque-specific cortical dimensions, a diversity of excitatory and inhibitory cell types with data-driven cell electrophysiology, data-driven population density and connectivity, detailed thalamic circuits, and realistic inputs from upstream structures such as cochlea and inferior colliculus. Additionally, we developed and included a local thalamic interneuron model that used extant pharmacological blockade data to delineate the source of intrinsic rhythmicity in this cell type. Overall, this model provides a quantitative theoretical framework for integrating and reproducing a wide range of experimental data from auditory circuits. Here we demonstrated this with respect to cortical oscillations, and highlighted how the model’s biological detail can be used to examine the origins of complex cortical activity.