Context-dependent neural processing of leg proprioception in Drosophila

Proprioception is critical for effective motor control: dysfunctions of the proprioceptive system can impair
balance, motor coordination, and motor learning. However, despite its importance, little is known about the initial
stages of proprioceptive processing in any animal, nor how this information is modulated by behavioral state. I
propose studying proprioception in the fly, D. melanogaster, whose proprioceptive system is more experimentally
accessible than that of vertebrates, but still analogous in its organization and function. I will combine experimental
and computational methods to study the flow of information from the proprioceptive sensory structure, the
femoral chordotonal organ, into genetically identifiable downstream circuits. In particular, I will characterize how
neural encoding changes during self vs. externally-generated movements, and how proprioceptive information
enters the brain to inform motor planning.
Test how perturbing specific inputs changes central encoding of imposed tibia movements. I will use patch-
clamp electrophysiology to record the activity of second-order proprioceptive neurons while moving the leg along
naturalistic and broadband, pseudo-random trajectories. I will then build a linear/nonlinear model to determine
the computations performed by each cell type. I will perturb inputs to central neurons and determine how these
perturbations alter neural encoding of leg movements.
Test the hypothesis that self- vs. externally-generated motions are differently encoded by some neurons. I
will record the activity of second-order neurons while the fly moves its leg. I will then replay those movements
and determine which neurons differently encode self- vs. externally-generated movements. I will characterize
how a neuron’s encoding changes and determine if there is an internal estimate of state expectations.
Determine how proprioceptive information entering the brain integrates with behavioral state and information
from other mechanoreceptors. Preliminary anatomical data suggests that a region of the brain, the wedge,
integrates multimodal mechanosensory cues from the legs and antennae. I will use 2-photon calcium imaging to
determine what proprioceptive information is relayed to this area and whether leg movement attenuates antennal
signals. I will then focus on how the central complex, a brain region important in motor planning, receives
proprioceptive input. I will use intracellular recordings and calcium imaging to ask which central complex neurons
encode proprioceptive information.
My long-term goal is to run my own research group focused on the function and evolution of the fly
proprioceptive system. Toward this end, my postdoctoral training is focused on the following goals: honing my
computational skills, developing management and mentoring skills, publishing and presenting my research, and
securing an independent, tenure-track position. I will be co-mentored by Drs. John Tuthill and Adrienne Fairhall
in the Physiology and Biophysics department at the University of Washington.

Grant Number: 5R00NS117657-05
NIH Institute/Center: NIH
Principal Investigator: Sweta Agrawal