Authors :
Presenting Author: Patrick Lawlor, MD, PhD – Children's Hospital of Philadelphia
Vijay Balasubramanian, PhD – Professor, Physics, Neuroscience, University of Pennsylvania; Ethan Goldberg, MD, PhD – Associate Professor, Neurology, Neuroscience, Children's Hospital of Philadelphia, University of Pennsylvania
Rationale:
Dravet Syndrome (DS) is a canonical childhood epilepsy syndrome associated with febrile seizures, treatment-resistant epilepsy, developmental delay/intellectual disability, autism spectrum disorder, and sudden unexplained death (SUDEP). DS is caused by variants in the gene
SCN1A, which encodes the voltage-gated sodium channel subunit Nav1.1, preferentially expressed in inhibitory interneurons implicated in preventing runaway circuit excitability (seizure). How the genetic abnormality underlying Dravet syndrome causes dysfunction at the single-neuron level is relatively well characterized. However, how single-neuron dysfunction leads to seizure – a circuit-level phenomenon – remains incompletely understood. Here, we use biologically-informed computational circuit simulations to reproduce the circuit-level phenotype (seizures). We then quantify circuit features which increase or decrease seizure probability.
Methods:
We model a two-dimensional cortical sheet of excitatory and inhibitory neurons using the Izhikevich mathematical model of single neurons. Synaptic connectivity was designed to be predominantly local. The population model was designed to be faithful to biological parameters such as cell type proportions, average firing rates, and input-output curves. The Izhikevich model of single neurons was augmented to produce inhibitory interneurons with Dravet-like pathophysiology, similar to what is seen in single-neuron electrophysiologic experiments (lower maximal firing rates with failure of action potential generation with high levels of synaptic or electrical stimulation). We initiated focal seizures with stimulation to a subset of the circuit, and monitored spread across the cortex (mimicking focal seizure onset with spread). We varied parameters of the circuit (such as synaptic connectivity strength) to investigate modulators of seizure probability.
Results:
Circuits containing Dravet-like inhibitory interneurons demonstrate a strong tendency to produce sustained, high firing rates in the excitatory neuron subpopulation outside of the stimulated area, mimicking focal seizure with spread. Circuits not containing Dravet-like inhibitory interneurons did not demonstrate this behavior. In the Dravet-like circuits, stronger excitatory-excitatory connectivity increased seizure probability, but stronger inhibitory-excitatory connectivity decreased seizure probability.
Conclusions:
Circuit-level simulations that incorporate a realistic, yet relatively simple, single-neuron model of Dravet syndrome reproduce the core phenotype - a strong tendency for seizure. This could be partially balanced by manipulating synaptic connectivity. This circuit-level model could be helpful for prototyping treatments and for exploring other important questions related to mechanisms of febrile seizures, learning, and development.
Funding:
NIH NINDS R25NS065745 to P.N.L. and R01NS110869 to E.M.G.