Long-range inputs and H-current regulate different modes of operation in a multiscale model of mouse M1 microcircuits

TitleLong-range inputs and H-current regulate different modes of operation in a multiscale model of mouse M1 microcircuits
Publication TypeConference Paper
Year of Publication2017
AuthorsDura-Bernal, S., Neymotin S. A., Suter B. A., Shepherd G. M., & Lytton WW.
Conference NameSociety for Neuroscience 2017 (SFN '17)
KeywordsSFN, Society for Neuroscience

We hypothesize the primary motor cortex (M1) may operate in different modes, or along a continuum of modes, characterized by the degree of activation of intratelencephalic (IT), corticothalamic (CT) and pyramidal tract (PT) neurons. The ``IT/CT-predominant'' mode would involve highly active IT and CT neurons but low PT activity; whereas the ``PT-predominant'' would exhibit the opposite pattern. In the ``IT/CT'' mode, a high level of H-current in PT neurons would modulate synaptic integration of inputs and reduce PT output activity. Downregulation of the H-current would lead to increased PT activity in the ``PT'' mode. The particular subset of M1 neurons targeted by different long-range inputs could also play a role in regulating the different modes: thalamic posterior nucleus (PO) and somatosensory cortical (S1 and S2) inputs would favor the ``IT/CT'' mode, whereas thalamic ventro-lateral (VL) and motor cortical (contralateral M1 and M2) would promote the ``PT'' mode.To study this hypothesis we developed a multiscale data-driven model of mouse M1 microcircuits, with over 10,000 cells distributed in a cylindrical volume of diameter 300 $μ$m and cortical depth 1350 $μ$m. L5 IT and PT neuron morphologies with 700+ compartments reproduced cell 3D reconstructions, and their ionic channel distributions were optimized within experimental constraints to reproduce in vitro recordings. The network includes over 30 million synaptic connections that depend on pre- and post-synaptic cell class and cortical depth. Data was based on optogenetic circuit mapping studies which determined that connection strengths vary within layer as a function of the neuron's cortical depth. The synaptic input distribution across cell dendritic trees – likely to subserve important neural coding functions – was also mapped using optogenetic methods and incorporated into the model. The network was driven by the main long-range inputs to M1: thalamus PO and VL, S1, S2, contralateral M1, M2, and orbital cortex (OC). We studied the effect on M1 of increased activity in each of these regions, and of different levels of H-current in PT neurons. Microcircuit dynamics and information flow were quantified using firing rates, oscillations, and information transfer measures (Granger causality and normalized transfer entropy). Preliminary results support the different modes of operation hypothesis; further exploration will help characterize the underlying mechanisms and determine interactions between the factors involved.