Neural Circuits for Error Correction

Introduction

To survive in natural habitats, animals move through space according to their goals. However, the uncertainties of the environment, alongside inevitable variations in neuromuscular signals, change the context in which a walking step occurs, leading to unintended movement. Thus, task performance can be jeopardized if the erroneous action is not rapidly corrected based on current posture and behavioral goals.

Research Gap

How these aspects of control functions are implemented and coordinated across the Central Nervous System remains unknown. Here we propose studying the circuits involved in monitoring our movements, since they are key intermediaries between motor planning and posture-dependent execution.

Fundamental Questions

Using the compact brain of the fly Drosophila melanogaster, we will ask two fundamental questions:

1. How is neural activity distributed across multiple networks integrated to estimate self-motion?
2. How is this internal estimate used to correct erroneous movement?

Methodology

Using a self-paced behavior, in which a fly drifting from a stable heading turns based on an internal drift estimate, we have found a circuit sensitive to angular velocity. This circuit is richly interconnected to the fly’s analogue of the spinal cord and higher-order brain areas, and is critical to drift-based turns.

Experimental Approach

We will leverage these results and combine them with electron microscopy, behavior, physiology, optogenetics, and modelling to study circuit mechanisms for course correction. We will:

1. Use connectomics and manipulations of neural activity to identify pathways involved in corrective turns.
2. Record from the identified neurons and correlate their activity with behavior.
3. Perturb cell type-specific neurons to test their role on self-motion computations and on corrective turns.
4. Test neural activity in different behavioral contexts.

Expected Outcomes

These experiments will establish unprecedented causal relationships between neural computations and movement and reveal the functional organization of distributed circuits for walking control.