Network effects of traumatic brain injury: from infra slow to high frequency oscillations

Abstract

AbstractTraumatic brain injury (TBI) can have a multitude of effects on neural functioning. In extreme cases, TBI can lead to seizures both immediately following the injury as well as persistent epilepsy over years to a lifetime. However, mechanisms of neural dysfunctioning after TBI remain poorly understood. To address these questions, we analyzed experimental data and developed a biophysical network model implementing effects of ion concentration dynamics and homeostatic synaptic plasticity to test effects of TBI on the brain network dynamics. We focus on three primary phenomena that have been reportedin vivoafter TBI: an increase in infra slow oscillations (<0.1 Hz), increase in delta (0.1 - 4 Hz) power, and the emergence of high frequency oscillations (HFOs) in the gamma range (30 - 100 Hz). We show that the infra slow oscillations can be directly attributed to extracellular potassium fluctuations, while the existence and characterization of HFOs is related to the increase in strength of synaptic weights from homeostatic synaptic scaling. The experimentally found transient increase in delta power can be attributed to the inter-HFO timings. We then show that buildup of high frequency oscillations in the injured region can lead to seizure-like events that span all neurons in the network; additional seizures can then be initiated in previously healthy regions. This study brings greater understanding of network effects of TBI, and how they can give rise to epileptic activity. This lays the foundation to begin investigating how injured networks can be healed and seizures prevented.Significance StatementThis project delineates and attempts to explain abnormalities seen in human brain following traumatic brain injury (TBI). TBI can lead to the development of seizures, which may last a lifetime and often become resistant to pharmaceutical treatments. The study identified key mechanisms responsible for occurrence of three characteristic changes in spatio-temporal network dynamics following TBI. This model provides predictions that can serve as a testing ground for potential therapeutic approaches.