Spike-dependent dynamic partitioning of the Locus coeruleus network through noradrenergic volume release in a simulation of nucleus core

Abstract

The Locus coeruleus (LC) modulates various neuronal circuits throughout the brain. Its unique architectural organization encompasses a net of axonal innervation that spans the entire brain; while its somatic core is highly compact. Recent research revealed an unexpected cellular input specificity within the nucleus that can give rise to various network states that either broadcast norepinephrine signals throughout the brain or pointedly modulate specific brain areas. Such adaptive input-output functions likely surpass our existing network models that build upon a given synaptic wiring configuration between neurons. As the distances between noradrenergic neurons in the core of the LC are unusually small, neighboring neurons could theoretically impact each other via volume transmission of NE. We therefore set out to investigate if such interaction could be mediated through noradrenergic alpha2-receptors in a spiking neuron model of the LC. We validate our model of LC neurons through comparison with experimental patch-clamp data and identify key variables that impact alpha2-mediated inhibition of neighboring LC neurons. Our simulation confirms a reliable autoinhibition of LC neurons after episodes of high neuronal activity that continue even after neuronal activity subsided. Also, dendro-somatic synapses inhibit spontaneous spiking in the somatic compartment of connected neurons in our model. We determined the exact position of hundreds of LC neurons in the mouse brain stem via a tissue clearing approach and, based on this, further determined that 25 percent of noradrenergic neurons have a neighboring LC neuron within less than a 25-micrometer radius. By modeling NE diffusion, we estimate that more than 15 percent of alpha2-adrenergic receptors fraction can bind NE within such diffusion radius. Our spiking neuron model of LC neurons predicts that repeated or long-lasting episodes of high neuronal activity induce partitioning of the gross LC network, and reduces the spike rate in neighboring neurons at distances smaller than 25 micrometers. As these volume-mediate neighboring effects are challenging to test with the current methodology, our findings can guide future experimental approaches to test this phenomenon and its physiological consequences.