Edge-based network analysis reveals frequency-specific network dynamics in aberrant anxiogenic processing in rats

Abstract

Uncovering interactions between edges of brain networks can reveal the organizational principle of the networks and also their dysregulations underlying aberrant behaviours such as in neuropsychiatric diseases. In this study, we looked into the applicability of edge-based network analysis in uncovering possible network mechanisms of aberrant anxiogenic processing. Utilizing a rat model of prodromal Parkinson’s disease we examined how a dorsomedial striatum–tied associative network (DSAN) may mediate context-based anxiogenic behaviour. Following dopamine depletion in the dorsomedial striatum, an exaggerated bottom-up signalling (posterior parietal-hippocampal-retrosplenial to anterior prefrontal-cingulate-amygdala regions) and gradient specific to the theta frequency in this network was observed. This change was accompanied by increased anxiety behaviour of the animals. By employing an edge-based approach in correlating informational flow (phase transfer entropy) with functional connectivity of all edges of this network, we further explore how the abnormal bottom-up signalling might be explained by alterations to the informational flow-connectivity motifs in the network. Our results demonstrate usage of edge-based network analysis in revealing concurrent informational processing and functional organization dynamics across multiple pathways in a brain network. This approach in unveiling network abnormalities and its impact on behavioural outcomes would be useful in probing the network basis of neuropsychiatric conditions.

Author Summary: This study employs a unique approach to study the neural basis of anxiety by quantifying the correlation of concurrent fluctuations in information flow and connectivity between all edges of the brain network involved. This is the first time that such an edge-based technique is used to analyse field potential signals of the brain in an animal model of a neuropsychiatric disorder. We demonstrate how such edge-based analysis complements other analyses such as local field potential power and gross network changes. This work helps lay the foundation for future investigations into decoding the network aberrations of a wide spectrum of neuropsychiatric disorders.