The ability to generate complex motor behaviors by assembling sequences of movements is essential for
purposeful actions and survival. Defects in the brain regions thought to drive such movement selection can lead
to behaviors becoming abnormally repetitive (e.g. autism spectrum disorder). Yet, the neural circuit architectures
that underlie this fundamental function of the nervous system remain poorly understood. A central model of a
neural circuit architecture that can account for how movements are assembled into sequences has emerged
from studies across multiple species. In this architecture, all movements are readied in parallel. Movements
within the sequence are then selected through hierarchical suppression, whereby earlier movements suppress
later ones. Prior studies have been unable to decipher how the model might arise from neuronal connectivity
and activity, in part due to the overwhelming complexity of neural circuitry in rodent models. We propose to
overcome this barrier through dissection of the neural circuitry underlying sequential body grooming movements
in the fruit fly, Drosophila melanogaster. Drosophila offer a useful compromise between complexity and
tractability as they display a rich behavioral repertoire, while their brains are numerically compact and have
uniquely identifiable neurons whose activity can be visualized and manipulated using powerful genetic-based
techniques. Grooming is ideal for probing the circuit principles of movement sequences because it consists of a
predictable sequence of distinct movements. Using this system, we previously showed that the Drosophila
grooming sequence has the hallmarks of a parallel model, and established an infrastructure of tools and
approaches to dissect the circuit basis of the model. The objective of this proposal is to define how the neural
circuit synaptic connectivity and activity ready the different movements in parallel and then produce hierarchical
suppression, two fundamental mechanisms predicted by the parallel model. In Aim 1, we will define how the
circuitry is organized to enable the movements to be readied in parallel. In Aim 2, we will elucidate how
hierarchical suppression controls grooming movement selection. These Aims will contribute to the first
description of a neural circuit architecture that produces sequential behavior via hierarchical suppression. Such
architectures are not only proposed to underlie movement sequences across species including humans, but can
also provide a general mechanism by which competing parallel inputs can be integrated to produce a prioritized
output. Thus, our proposed study in the fruit fly will be relevant to other animals, both for understanding how
complex motor behaviors are produced and for understanding neural circuit organization and function more
broadly.
