Generative modelling of the thalamo-cortical circuit mechanisms underlying the neurophysiological effects of ketamine
Alexander D Shaw, Suresh D Muthukumaraswamy, Neeraj Saxena, Rachael L Sumner, Natalie Adams, Rosalyn J Moran, Krish D Singh
bioRxiv Preprint Server May 5, 2020 preprint DOI: 10.1101/688044 via bioRxiv
Summary
AI-generated from the abstractKetamine alters brain oscillations, increasing high-frequency gamma waves and reducing low-frequency alpha and theta waves. A thalamo-cortical model better explained these changes than a cortex-only model. The model showed that ketamine increases specific synaptic connections: from superficial pyramidal cells to inhibitory interneurons via AMPA and NMDA receptors, and within-layer-5 pyramidal cell gain control via GABA-A and NMDA receptors. Receptor time-constants remained unchanged. These findings support using generative models to understand oscillatory data and provide computational evidence that ketamine alters local neural coupling through multiple neurotransmitter systems.
Study at a glance
| Characteristics | Computational modeling study with Bayesian model selection |
|---|---|
| Intervention | Ketamine |
| Dose | subanaesthetic |
| Topics | Ketamine |
| Keywords | Ketamine effects Ketamine mechanisms Computational neuroscience Brain modeling Neural modeling |
| Citations | 2 |
| Key finding | Ketamine-induced changes in alpha and gamma oscillations are explained by increased synaptic connectivity in specific thalamo-cortical circuits involving AMPA, NMDA, and GABA-A receptors. |
Abstract
Cortical recordings of task-induced oscillations following subanaesthetic ketamine administration demonstrate alterations in amplitude, including increases at high-frequencies (gamma) and reductions at low frequencies (theta, alpha). To investigate the population-level interactions underlying these changes, we implemented a thalamo-cortical model (TCM) capable of recapitulating broadband spectral responses. Compared with an existing cortex-only 4-population model, Bayesian Model Selection preferred the TCM. The model was able to accurately and significantly recapitulate ketamine-induced reductions in alpha amplitude and increases in gamma amplitude. Parameter analysis revealed no change in receptor time-constants but significant increases in select synaptic connectivity with ketamine. Significantly increased connections included both AMPA and NMDA mediated connections from layer 2/3 superficial pyramidal cells to inhibitory interneurons and both GABAA and NMDA mediated within-population gain control of layer 5 pyramidal cells. These results support the use of extended generative models for explaining oscillatory data and provide in silico support for ketamine’s ability to alter local coupling mediated by NMDA, AMPA and GABA-A.