Neural circuits and motor sequences

Fly grooming is composed of discrete motor programs selected in sequence

Our characterization of normal fly grooming behavior shows modularity and progression.  Grooming contains separable motor programs: the front legs clean the eyes, antennae, and head, while the rear legs sweep the abdomen, wings and thorax.  Human observers identify these discrete subroutines, and different groups of neurons control them.  Flies usually clean their bodies in an anterior to posterior progression, starting from their heads and continuing to their abdomen, wings, and thorax, but the sequence is not rigidly stereotyped.  Rather, the sequence emerges because the probability of performing each behavior changes with time and dust distribution on the body.  This suggests that a suppression hierarchy, rather than activation chain or delay line algorithms, produces the sequence.  This mechanism may be a general solution to selecting among competing behaviors in a range of contexts.  We use classical ethology and genetic manipulation of neuronal activity to demonstrate that a suppression hierarchy governs the probabilistic progression of grooming behavior (Seeds et al 2014 eLife).  

A multi-layered command circuit controls antennal grooming

What neurons and circuits control each grooming subroutine, and how are these circuits selected in a mutually exclusive way?  Using the genetic reagents identified in our initial behavior screens, we generated intersectional tools to determine which neurons could specifically evoke antennal grooming.  We show that several distinct types of neurons form a functionally connected circuit to coordinate this motor program.  Although very few neurons are involved, they are highly interconnected and form recognizable circuit motifs.  ​We will now explore the computations these motifs enable. (Hampel et al 2015 eLife in revision).

Automatic detection of grooming movements reveals syntax and dynamics

Behavioral quantification can be performed manually, but automatic detection of grooming movements enables discovery of subtle patterns in transition probabilities.  In collaboration with Kristin Branson (Janelia), we have developed a method to score grooming movements from video.  We are analyzing spontaneous and dust-induced grooming to search for internal sequence triggers and exploring the nature of the grooming sequence in decapitated flies to define the role the brain plays in ordering the sequence progression (Ravbar et al in preparation).

FUTURE DIRECTIONS

BEHAVIOR: We will use higher resolution imaging and manipulation of grooming behavior to explore sensory cues and neuronal computations that instigate behavior mode transitions.

CIRCUITS: We are conducting genetic screens for neurons specifically affecting grooming sequence progression to get at precise neural circuits that establish hierarchy, suppress competing behaviors, and enable choice to change with time or sensory information.

ACTIVITY: Ultimately we have to understand the patterns of electrical activity in relevant neurons embedded in networks in the biomechanical structure of the fly.  We are developing a preparation to perform two-photon calcium imaging in the ventral nerve cord during grooming behavior.

History

In its previous location (HHMI Janelia Research Campus 2006-2015), the Simpson Lab has worked on the neural circuits that mediate hunger (Albin et al Current Biology 2015) and an atlas of motor neurons that control proboscis extension (McKellar et al, in preparation).  As part of the Fly Olympiad, we have examined the neurons required for male fertility (Chen et al, in preparation) and the behavioral flexibility of courtship song (Britton et al, in preparation).  We have an on-going commitment to collaborative projects and interest in tool development for better ways to manipulate the nervous system of the fly.