Owen Mackwood: Homeostatic and functional implications of interneuron plasticity

BCCN Berlin

Preserving the functional capacities of the brain despite ongoing changes, both inside the organism and out in the world, necessitates a multitude of homeostatic mechanisms. For neuronal networks within the brain, inhibitory neurons play a central role in computation, and are likely involved in maintaining its homeostasis. However, it remains an open question what mechanisms cause inhibitory neurons to develop their observed properties, and subsequently preserve them. This thesis therefore aims to determine the homeostatic capabilities of inhibitory neurons and elucidate the resulting functional consequences, using mathematical and computational techniques.

The central hypothesis of this thesis is that some inhibitory interneurons slowly modulate their firing rates so as to maintain the long-term activity of excitatory neurons at a homeostatic set-point. Consequently, we begin by deriving a plasticity rule that minimizes network-wide deviations from that set-point, by regulating the activity of interneurons. Although the rule is not biologically plausible, we propose two approximations to it that are. Those rules make each interneuron responsive to the excitatory population it inhibits, with one of them relying on retroaxonal signalling, and the other relying on diffusion-like signalling. They both exhibit comparable, though distinct, homeostatic capabilities. To contrast with those normative rules, we characterise the homeostatic properties of several other rules which alter the activity of an interneuron when the neurons that drive it deviate from the set-point. These rules induce a competition between excitatory neurons, causing network activity to become sparse.

In the second part of this thesis, we investigate how our retroaxonal rule affects functional properties of sensory cortex. We demonstrate that it can account for several experimentally reported results, including co-tuning of excitatory and inhibitory currents, and the development of excitatory-inhibitory cell assemblies. In summation, this thesis provides new insight into how regulating interneuron activity can be homeostatic for neuronal networks, and reveals the potential implications for development and preservation of brain function.

Additional Information

PhD defense

Organizer

Henning Sprekeler