What can self-connected neurons do? A unifying theory of autaptic neural function in invertebrates and mammals

The neuron is the fundamental building block in the nervous systems of all animal species, from C. elegans with its 302 neurons, to humans with their highly evolved cortex alone containing more than 10 billion neurons. The action potential, or spike, is the basic unit of neural signalling.

Most neurons communicate through chemical synapses, which project from the axon of one neuron to the cell body or dendrites of another. An autapse is a synapse connecting the neuron to itself. In this project, we will use mathematical modelling to achieve two objectives: 1. Investigate the different roles of autapses in neural processing as observed experimentally. 2.

Explore autapses as a potential route to chaos and its role in healthy and pathological neural function.

Objective 1: The role of autapses in neural processing: Neurons are either excitatory, where presynaptic spikes increase the likelihood for postsynaptic spikes, or inhibitory, where presynaptic spikes suppress postsynaptic spikes.

Autapses are present in the nervous systems of many species, but their functional role in neural processing and computations is not resolved.

In the invertebrate Aplysia, excitatory autaptic motor neurons maintain feeding behaviour by promoting persistent activity.

In humans and rodents, excitatory prefrontal autapses enhance bursting dynamics, which may boost neural information transmission, but do not enhance persistent activity as in the Aplysia. The reason for this difference is unknown.

Inhibitory autaptic neurons in humans and rodents can modulate excitability in many ways, depending on their place within cortex.

Uncovering the origin of this variability would benefit from a unifying approach that effectively correlates the properties of different autaptic types to the behaviours they have evolved to support. As a first step toward achieving this objective, we developed a new general model of an autaptic neuron.

Different parameter ranges in our model correspond to distinct autaptic types, which will be explored further in this project through model simulations and mathematical analysis based on methods from dynamical systems theory. We will also investigate how different autapses influence network dynamics leading to distinct behaviours.

Objective 2: Autapses as a potential route to chaos in prefrontal neurons: Chaos refers to complex dynamics in which small variations in the current state of a system can lead to large deviations in its future behaviour.

Emerging research is increasingly emphasising the pivotal role of chaotic dynamics in understanding the complexities of brain function.

Chaotic dynamics have been associated with the transition between sleep and wakefulness, efficient cognitive search, flexibility, and creativity, as well as with cognitive decline in older age.

On the other hand, excessive chaos has been linked to attentional deficits in ADHD, disorganised thoughts in schizophrenia, and mood variations in bipolar disorder.

Models of chaotic dynamics have also been employed to understand epilepsy, Alzheimer's, Parkinson's, Huntington's, and hypokinesia.

Motivated by these diverse implications, our project aims to delve into the fundamental neuronal properties that give rise to such phenomena. To achieve this, we will employ dynamical systems modelling and analysis as our primary methodologies.

Preliminary research has indicated that the proposed autaptic model neuron can exhibit chaos within excitatory prefrontal parameter ranges.

Given the central role of the prefrontal cortex in supporting the proper function of higher cognitive abilities and its implications in various neuropathologies, it is important to investigate how excitatory prefrontal autapses may influence network dynamics.

The insights from our project could offer a nuanced understanding of neural function in both health and disease, thereby informing future diagnostics and treatments.