Motor cortical dynamics are shaped by multiple distinct subspaces during naturalistic behavior

Abstract

Behavior relies on continuous influx of sensory information about the body and the environment. In primates, cortex integrates somatic feedback to accurately reach and manipulate objects. Yet, in many experimental regimes motor cortex seems paradoxically to operate as a feedforward, rather than feedback-driven, system. Here, we recorded simultaneously from motor and somatosensory cortex as monkeys performed a naturalistic reaching and object interaction behavior. We studied how unexpected feedback from behavioral errors influences cortical dynamics. Motor cortex generally exhibited robust feedforward dynamics, yet displayed feedback-driven dynamics surrounding correction of behavioral errors. We then decomposed motor cortical activity into orthogonal subspaces capturing communication with somatosensory cortex or behavior-generating dynamics. During error correction, the communication subspace became feedback-driven, while the behavioral subspace maintained feedforward dynamics. We therefore demonstrate that cortical activity is compartmentalized within distinct subspaces that shape the population dynamics, enabling flexible integration of salient inputs with ongoing activity for robust behavior.

Motor cortex activity across movement speeds is predicted by network-level strategies for generating muscle activity

“…For example, this is true in primary somatosensory cortex during cycling 4 and (unsurprisingly) in primary visual cortex 4,33 . Tangling will also be high when unpredictable external inputs dominate the motor cortex response 34 . For example, autonomous dynamics provide a poorer fit to the data during grasping 35 .…”

Rotational dynamics in motor cortex are consistent with a feedback controller

“…Recent methods that exploit simultaneous recordings from multiple brain areas provide a promising tool to identify input signals to a given circuit (Kohn et al, 2020; Perich et al, 2018; Semedo et al, 2019). Using these techniques, Perich et al, (2020) provides evidence of a communication subspace between the somatosensory and motor cortices that contributes to a substantial amount of the variance in MC, consistent with inputs playing a key role in motor cortical dynamics. Studies that perturb neural circuits through stimulation, cooling probes or optogenetics will also provide valuable insight into how inputs are transformed by MC (Guo et al, 2020; Hore et al, 1977; Li et al, 2016; Nashef et al, 2018, 2019; Perich et al, 2020; Svoboda and Li, 2018; Takei et al, 2020).…”

“…Using these techniques, Perich et al, (2020) provides evidence of a communication subspace between the somatosensory and motor cortices that contributes to a substantial amount of the variance in MC, consistent with inputs playing a key role in motor cortical dynamics. Studies that perturb neural circuits through stimulation, cooling probes or optogenetics will also provide valuable insight into how inputs are transformed by MC (Guo et al, 2020; Hore et al, 1977; Li et al, 2016; Nashef et al, 2018, 2019; Perich et al, 2020; Svoboda and Li, 2018; Takei et al, 2020). For example, deactivating cerebellar output can substantially impact preparatory activity in the MC and feedback responses to mechanical loads (Chabrol et al, 2019; Conrad et al, 1974; Gao et al, 2018; Meyer-Lohmann et al, 1975).…”

Inferring brain-wide interactions using data-constrained recurrent neural network models

“…Estimating population-wide inputs at this scale produces an unprecedented view into multi-region interactions. Furthermore, CURBD scales readily beyond two interconnected regions to brain-wide interactions 19 , circumventing a limitation of many existing approaches [23][24][25][26] . In a single-layer network, each source unit connects to a target unit with a directed interaction weight given by the vector J.…”