Songs, Sequences and Time: Neural mechanisms of sequence generation in the brain.

Michale Fee

Bell Labs Lucent Technology

Sensitivity to temporal order is a nearly universal aspect of brain function - at the sensory, motor, and cognitive levels. The ability of the brain to step rapidly through a learned sequence of states underlies not only the performance of complex motor tasks such as speech, but perhaps our ability to think and plan as well. While little is known about the neural and circuit mechanisms underlying the generation and learning of sequences, songbirds provide a marvelous animal model in which to study these phenomena.

We have used a new miniature motorized microdrive to record from large numbers of neurons in two premotor song-control brain areas in the songbird: nucleus RA, which projects to motor neurons of the vocal organ, and nucleus HVC, which projects to RA. During singing, each RA neuron generates a distinctive and reproducible sequence of spike bursts. Until recently, a widely held view was that the burst patterns in RA are largely generated by circuitry intrinsic to RA, an idea we have directly tested by measuring inputs to RA from nucleus HVC. We found that burst sequences in RA are directly driven from HVC, and that each antidromically identified RA-projecting HVC neuron generates a single brief burst of spikes at one moment in the song. We suggest that HVC neurons produce a detailed code for the temporal ordering of vocal gentures on a 10 ms timescale, and form an explicit representation of time in the brain. Such a sparse code for temporal order has broad implications for sequence learning.