Diversity of neuronal activity is provided by hybrid synapses

TL;DR

This paper models neural networks with hybrid synapses to explore how electrical and chemical connections influence diverse neural activity states, revealing that mixed synapses enable various synchronized and ripple dynamics, crucial for understanding neural function.

Contribution

It introduces a comprehensive neural network model with hybrid synapses, analyzing how electrical and chemical connections jointly produce diverse dynamical states and self-organization.

Findings

Electrical coupling controls synchrony in balanced networks.

Strong electrical synapses lead to network-wide synchronization.

Small changes in chemical weights induce various dynamical states.

Abstract

Many experiments have evidenced that electrical and chemical synapses -- hybrid synapses -- coexist in most organisms and brain structures. The role of electrical and chemical synapse connection in diversity of neural activity generation has been investigated separately in networks of varying complexities. Nevertheless, theoretical understanding of hybrid synapses in diverse dynamical states of neural networks for self-organization and robustness still has not been fully studied. Here, we present a model of neural network built with hybrid synapses to investigate the emergence of global and collective dynamics states. This neural networks consists of excitatory and inhibitory population interacting together. The excitatory population is connected by excitatory synapses in small world topology and its adjacent neurons are also connected by gap junctions. The inhibitory population is only connected by chemical inhibitory synapses with all-to-all interaction. Our numerical simulations show that in the balanced networks with absence of electrical coupling, the synchrony states generated by this architecture are mainly controlled by heterogeneity among neurons and the balance of its excitatory and inhibitory inputs. In balanced networks with strong electrical coupling, several dynamical states arise from different combinations of excitatory and inhibitory weights. More importantly, we find that these states, such as synchronous firing, cluster synchrony, and various ripples events, emerge by slight modification of chemical coupling weights. For large enough electrical synapse coupling, the whole neural networks become synchronized. Our results pave a way in the study of the dynamical mechanisms and computational significance of the contribution of mixed synapse in the neural functions.