UNVEILING THE NEURAL NETWORK UNDERLYING THE GENERATION OF NEURAL SEQUENCES IN THE HVC THROUGH COMPUTATIONAL MODELING

Abstract

Birdsong offers a unique model system to understand how a developing brain – once given a set of purely acoustic targets – teaches itself the vocal-tract gestures necessary to imitate those sounds. Like human infants, juvenile male zebra finches (Taeniopygia guttata) passes through the stages of learning the vocal-motor gestures of adult sounds.

The HVC nucleus (avian brain region, used as a proper name) is a cortical nucleus in the forebrain that is responsible for the songbird’s singing as well as the learning process of his song. The HVC consists of three neural populations: basal-ganglia-projecting (HVCX) neurons, forebrain-projecting (HVCRA) neurons and interneurons (HVCINT). Each neuron population has its own cellular, electrophysiological and functional properties. In particular, HVCRA neurons emit a single 6-10 ms burst of action potentials at the same exact time during each rendition of song, HVCX neurons fire 1 to 3 bursts that are also time locked to vocalizations, while HVCINT neurons fire and burst randomly with high firing frequency throughout song with no significant pattern. As a population, these three classes of HVC neurons form an explicit representation of time and are responsible for orchestrating song learning and production; yet little is known about their functional connectivity within nucleus HVC and how they work cooperatively to control learning and singing.

Very few mathematical models have been developed to describe HVC’s neural activity, and all of the generated models were either non-biological plausible or replicating in vitro data collected from brain slices. We developed a conductance-based Hodgkin Huxley model for the three classes of HVC neurons and connected them in several networks via different architecture patterning scenarios with the aim to replicate the in vivo firing patterning behaviors. We are able throughout these networks to reproduce the in vivo behavior of these neurons as shown by the experimental recordings. The study of the synaptic architecture of the HVC nucleus has given us insights on the nature of the physiological changes taking place inside the bird’s brain during learning and vocalization providing a large step towards biologically plausible descriptions of the underlying in vivo neural networks.